INDEX
Explanations
instances of the article "a" and its variations in the text
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
50
+0.45
2.3%
1967
+0.15
0.8%
1896
+0.12
0.6%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1984
+0.45
0.11
1896
+0.15
0.08
1034
+0.12
0.07
Negative Logits
<bos>
-2.19
//};
-0.66
};*/
-0.61
//----
-0.61
//---
-0.60
confiable
-0.60
//});
-0.59
automáticamente
-0.58
///**
-0.56
ⓧ
-0.55
POSITIVE LOGITS
Haci
0.91
viciss
0.86
Compañ
0.84
valencia
0.79
quoique
0.79
véhic
0.79
conflic
0.76
sappi
0.76
hacienda
0.75
Áng
0.74
Activations Density 0.485%