INDEX
Explanations
Quotation marks and dialogue within the text
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
50
+0.29
1.3%
381
+0.17
0.8%
2019
+0.16
0.7%
Correlated Neurons
Index
P. Corr.
Cos Sim.
545
+0.29
0.08
82
+0.17
0.07
270
+0.16
0.07
Negative Logits
<bos>
-2.31
ⓧ
-0.90
springfox
-0.82
дописавши
-0.73
<?
-0.73
EndGlobalSection
-0.68
/***
-0.67
Roskov
-0.67
FunctionFlags
-0.66
EndProject
-0.65
POSITIVE LOGITS
unspeak
1.51
reluct
1.48
apprehen
1.44
disagre
1.43
increa
1.39
accla
1.36
tolerably
1.33
affor
1.33
gaily
1.31
maneu
1.31
Activations Density 0.202%