INDEX
Explanations
the word "Matter" in legal or formal contexts
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
492
+0.13
0.7%
376
+0.12
0.7%
463
+0.10
0.6%
Correlated Neurons
Index
P. Corr.
Cos Sim.
370
+0.13
0.01
505
+0.12
0.01
492
+0.10
0.01
Negative Logits
·
-2.22
¯
-2.09
Ĩ
-1.94
JUD
-1.77
ĥ
-1.76
®
-1.71
µ
-1.63
¿½
-1.63
İ
-1.59
ĥ½
-1.58
POSITIVE LOGITS
pool
1.87
nets
1.72
ingham
1.66
horiz
1.66
houses
1.66
ière
1.63
ÃŃvel
1.60
table
1.59
bugs
1.55
moeten
1.55
Activations Density 0.018%