INDEX
Explanations
sentences or phrases related to legal issues and statistics
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
2034
+0.13
0.4%
1535
+0.11
0.3%
919
+0.10
0.3%
Correlated Neurons
Index
P. Corr.
Cos Sim.
382
+0.13
0.09
22
+0.11
0.07
310
+0.10
0.04
Negative Logits
gsx
-0.93
vinci
-0.91
fordable
-0.84
rodriguez
-0.84
uncin
-0.84
carina
-0.80
pajero
-0.78
dolom
-0.76
emphat
-0.76
nece
-0.74
POSITIVE LOGITS
Nor
0.74
Therefore
0.74
nor
0.72
Nor
0.68
Instead
0.66
Hence
0.63
Consequently
0.63
Therefore
0.63
Thus
0.62
therefore
0.62
Activations Density 0.595%