INDEX
Explanations
questions and answers in interview segments
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
1150
+0.10
0.3%
655
+0.08
0.2%
1403
+0.08
0.2%
Correlated Neurons
Index
P. Corr.
Cos Sim.
59
+0.10
0.03
1314
+0.08
0.03
1973
+0.08
0.02
Negative Logits
autunno
-0.62
sento
-0.60
Giugno
-0.59
haer
-0.59
depositphotos
-0.58
apparti
-0.58
sulphuric
-0.57
haviour
-0.56
Chapitre
-0.56
veau
-0.55
POSITIVE LOGITS
questions
0.89
Questions
0.82
questions
0.77
Answers
0.75
answers
0.74
Q
0.74
Q
0.73
answered
0.70
Questions
0.70
question
0.67
Activations Density 0.133%