INDEX
Explanations
This neuron detects the section heading “OPINION AND ORDER” in court-opinion documents.
New Auto-Interp
Negative Logits
Shift
-0.07
Broken
-0.07
ueur
-0.07
Jim
-0.07
липня
-0.07
STRU
-0.06
Spaces
-0.06
.permissions
-0.06
Jewel
-0.06
Wars
-0.06
POSITIVE LOGITS
_ver
0.06
_AST
0.06
insurgency
0.06
ssid
0.06
Replacement
0.06
<Car
0.06
,<
0.06
Grass
0.06
achable
0.06
/**/*.
0.06
Activations Density 0.003%