INDEX
Explanations
technical terms related to electronic devices and communication technologies
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
1967
+0.26
0.9%
674
+0.14
0.5%
1081
+0.13
0.4%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1081
+0.26
0.06
1597
+0.14
0.03
1363
+0.13
0.04
Negative Logits
effe
-1.88
secon
-1.83
inev
-1.78
madonna
-1.77
pessi
-1.75
squa
-1.75
wien
-1.75
mef
-1.73
igno
-1.73
inder
-1.73
POSITIVE LOGITS
розта
0.71
сьогодні
0.68
примі
0.66
відді
0.65
допомо
0.64
краще
0.64
близь
0.63
fillType
0.62
tisgarh
0.62
варто
0.62
Activations Density 0.449%