INDEX
Explanations
phrases related to physical fitness, exercise, and health
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
50
+0.17
1.0%
1262
+0.11
0.6%
1491
+0.09
0.5%
Correlated Neurons
Index
P. Corr.
Cos Sim.
75
+0.17
0.03
1262
+0.11
0.02
1331
+0.09
0.02
Negative Logits
<bos>
-3.09
/***
-0.83
//---
-0.68
/*!
-0.61
<?
-0.57
public
-0.57
rehabilitate
-0.56
adopt
-0.55
<>
-0.54
ⓧ
-0.54
POSITIVE LOGITS
stockholm
1.24
Minang
1.23
maroc
1.23
Juf
1.19
maneu
1.19
jaya
1.16
effe
1.16
Balance
1.15
équilibr
1.13
foon
1.11
Activations Density 0.084%