INDEX
Explanations
mentions of the word "visit"
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
871
+0.13
0.7%
528
+0.11
0.6%
1047
+0.11
0.6%
Correlated Neurons
Index
P. Corr.
Cos Sim.
871
+0.13
0.02
597
+0.11
0.02
889
+0.11
0.02
Negative Logits
<bos>
-1.86
<?
-0.77
ⓧ
-0.76
-0.72
<?
-0.65
/**
-0.64
prek
-0.62
/***
-0.62
osal
-0.59
iria
-0.58
POSITIVE LOGITS
Vis
1.35
Vis
1.30
vis
1.26
VIS
1.21
vis
1.16
visibility
1.11
VIS
1.09
Visibility
1.08
visible
0.94
Visible
0.92
Activations Density 0.256%