INDEX
Explanations
references to the term "elite" and related concepts
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
50
+0.29
1.4%
555
+0.13
0.7%
596
+0.13
0.6%
Correlated Neurons
Index
P. Corr.
Cos Sim.
596
+0.29
0.02
555
+0.13
0.01
1872
+0.13
0.01
Negative Logits
<bos>
-2.15
lateinit
-0.61
AddWithValue
-0.59
HasAnnotation
-0.58
AutoScaleMode
-0.56
和田
-0.55
rungsseite
-0.54
//---
-0.53
/***
-0.53
glPushMatrix
-0.53
POSITIVE LOGITS
Elite
1.16
Elite
1.11
elite
1.03
jaya
1.02
Breslau
1.00
🤣🤣
0.96
kyo
0.95
Schrö
0.93
élite
0.93
jati
0.93
Activations Density 0.069%