INDEX
Explanations
the word "the" occurring in the text
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
50
+0.29
1.5%
1363
+0.10
0.5%
1331
+0.10
0.5%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1967
+0.29
0.05
1325
+0.10
0.04
1516
+0.10
0.04
Negative Logits
<bos>
-2.59
ⓧ
-0.93
/**
-0.92
<?
-0.85
-0.79
/***
-0.76
<?
-0.69
///**
-0.67
Transkript
-0.66
Transcripción
-0.66
POSITIVE LOGITS
magis
0.96
italia
0.91
particolar
0.88
susun
0.85
mezza
0.84
gamba
0.83
seksi
0.82
palab
0.81
tramont
0.81
padang
0.81
Activations Density 0.393%