INDEX
Explanations
descriptions of food that convey lightness and freshness
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
50
+0.27
1.1%
1677
+0.12
0.5%
1805
+0.10
0.4%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1677
+0.27
0.03
1805
+0.12
0.02
1376
+0.10
0.02
Negative Logits
<bos>
-1.81
twimg
-0.72
glTranslatef
-0.58
glBegin
-0.56
glTexCoord
-0.56
HttpDelete
-0.55
expel
-0.55
redefine
-0.54
HasIndex
-0.54
<?
-0.54
POSITIVE LOGITS
Minang
1.12
light
1.02
soulign
0.97
Light
0.96
Light
0.94
Lightfoot
0.94
lightness
0.94
Putih
0.94
light
0.93
quoique
0.91
Activations Density 0.041%