INDEX
Explanations
names related to organized crime families
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
538
+0.16
0.7%
1392
+0.13
0.6%
196
+0.13
0.5%
Correlated Neurons
Index
P. Corr.
Cos Sim.
538
+0.16
0.03
196
+0.13
0.02
1392
+0.13
0.02
Negative Logits
testify
-0.46
uxxxx
-0.45
ValueStyle
-0.45
centerTitle
-0.45
hintText
-0.44
Duisburg
-0.42
dataclass
-0.42
ensement
-0.42
Nazionale
-0.41
leqq
-0.41
POSITIVE LOGITS
Bor
1.50
bor
1.32
Bor
1.32
BOR
1.32
BOR
1.12
bor
1.11
Borrow
1.00
borrow
0.99
Borrow
0.93
borrowing
0.91
Activations Density 0.108%