INDEX
Explanations
terms related to altitude and elevation measurements
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
50
+0.25
1.0%
1385
+0.19
0.8%
906
+0.14
0.6%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1385
+0.25
0.09
906
+0.19
0.01
1013
+0.14
0.06
Negative Logits
<bos>
-1.99
ⓧ
-0.98
/**
-0.89
-0.78
/*
-0.75
Obrázky
-0.72
dras
-0.68
<?
-0.67
sizeCache
-0.65
<?
-0.64
POSITIVE LOGITS
pandan
1.12
swarovski
1.09
getHeight
1.03
malheureux
1.03
despotism
1.01
impractica
1.01
genshin
0.99
ecru
0.98
Bartholo
0.97
mépris
0.97
Activations Density 0.404%