INDEX
Explanations
references to registration, forums, and online communities
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
50
+0.23
1.3%
1103
+0.12
0.7%
82
+0.09
0.5%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1103
+0.23
0.06
1516
+0.12
0.06
82
+0.09
0.05
Negative Logits
<bos>
-2.56
ⓧ
-1.00
-0.98
<?
-0.94
/***
-0.86
///**
-0.83
/**
-0.82
<?
-0.81
springfox
-0.70
/*
-0.68
POSITIVE LOGITS
|
0.75
|
0.75
('|0.72
véhic
0.71
("|0.70
particolar
0.69
abello
0.69
$|
0.68
bicic
0.68
"|
0.68
Activations Density 0.084%