INDEX
Explanations
Legal texts
This neuron detects the names of judges (or other proper names) listed in the court’s panel heading of judicial opinions.
New Auto-Interp
Negative Logits
serialization
-0.08
Shoot
-0.07
ships
-0.07
COMP
-0.06
phải
-0.06
NIH
-0.06
figure
-0.06
ิก
-0.06
ประเทศ
-0.06
MMP
-0.06
POSITIVE LOGITS
хозя
0.07
>()↵↵
0.06
breeze
0.06
,存于
0.06
financial
0.06
Blues
0.06
.co
0.06
life
0.06
Ty
0.06
aws
0.06
Activations Density 0.016%