INDEX
Explanations
words related to physical touch or interaction
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
1482
+0.09
0.3%
50
+0.08
0.3%
1350
+0.07
0.3%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1002
+0.09
0.05
1748
+0.08
0.04
620
+0.07
0.04
Negative Logits
<bos>
-1.21
public
-0.62
///**
-0.62
hline
-0.61
ⓧ
-0.60
//
-0.60
<tfoot>
-0.59
onPause
-0.58
enumerate
-0.58
const
-0.58
POSITIVE LOGITS
touch
2.23
TOUCH
2.16
Touch
2.15
touch
2.11
Touch
1.97
touches
1.89
touching
1.76
touched
1.76
TOUCH
1.73
meis
1.70
Activations Density 0.142%