INDEX
Explanations
Spanish words and names
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
971
+0.16
0.6%
1967
+0.15
0.6%
1343
+0.14
0.5%
Correlated Neurons
Index
P. Corr.
Cos Sim.
981
+0.16
0.09
227
+0.15
0.08
1597
+0.14
0.04
Negative Logits
biles
-0.55
poliuret
-0.52
tempes
-0.52
števil
-0.51
bici
-0.51
poliester
-0.50
dentes
-0.48
tificial
-0.46
nori
-0.46
braccia
-0.45
POSITIVE LOGITS
WebElementEntity
0.65
Outre
0.65
vété
0.64
compréhen
0.64
indestru
0.64
vêtement
0.63
Souha
0.63
Autre
0.62
Prede
0.62
éclairage
0.62
Activations Density 0.416%