INDEX
Explanations
This neuron detects the “Court of Common Pleas” heading (i.e. jurisdiction labels in legal case captions).
New Auto-Interp
Negative Logits
interpreted
-0.07
migrated
-0.06
ftar
-0.06
VP
-0.06
Sadd
-0.06
Fan
-0.06
currentDate
-0.06
vydání
-0.06
-side
-0.06
ektör
-0.06
POSITIVE LOGITS
wages
0.07
_deps
0.07
cubes
0.07
stationary
0.06
التو
0.06
Expense
0.06
_spec
0.06
/functions
0.06
Format
0.06
iple
0.06
Activations Density 0.002%