INDEX
Explanations
terms related to legal and political scenarios, particularly focused on revolutionary themes and governmental actions
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
674
+0.13
0.4%
755
+0.12
0.4%
1036
+0.10
0.3%
Correlated Neurons
Index
P. Corr.
Cos Sim.
755
+0.13
0.03
971
+0.12
0.02
1008
+0.10
0.02
Negative Logits
étrang
-1.02
volet
-1.01
levier
-0.99
jouet
-0.91
malheureux
-0.90
aveug
-0.89
clôture
-0.88
héro
-0.87
strick
-0.87
exemplaire
-0.86
POSITIVE LOGITS
<bos>
0.69
comuna
0.69
Inoltre
0.66
ideolog
0.66
excelente
0.66
such
0.66
such
0.66
Such
0.63
Ottimo
0.63
Explicación
0.63
Activations Density 0.075%