INDEX
Explanations
information related to magazine covers
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
50
+0.16
0.8%
1096
+0.11
0.5%
1413
+0.10
0.5%
Correlated Neurons
Index
P. Corr.
Cos Sim.
568
+0.16
0.03
1096
+0.11
0.03
90
+0.10
0.03
Negative Logits
<bos>
-2.94
/***
-0.99
ⓧ
-0.99
///**
-0.87
ánchez
-0.72
-0.69
//@
-0.65
lateinit
-0.64
<>
-0.64
/*!
-0.63
POSITIVE LOGITS
lele
1.37
Minang
1.24
Karang
1.24
kasa
1.21
fua
1.18
bandung
1.17
jaya
1.15
kuli
1.15
saar
1.14
kase
1.11
Activations Density 0.084%