INDEX
Explanations
references to political figures and events
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
1577
+0.23
0.8%
394
+0.14
0.5%
50
+0.12
0.4%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1001
+0.23
0.14
81
+0.14
0.04
344
+0.12
0.09
Negative Logits
swarovski
-1.31
unwarran
-1.24
ecru
-1.24
increa
-1.24
scrat
-1.21
reluct
-1.15
ftu
-1.15
milf
-1.14
hairc
-1.14
disagre
-1.13
POSITIVE LOGITS
political
0.68
Demokrat
0.65
Political
0.63
presidential
0.62
voters
0.62
partisan
0.58
electoral
0.56
election
0.56
elections
0.56
Republi
0.55
Activations Density 8.207%