INDEX
Explanations
direct quotations or excerpts from sources
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
2019
+0.11
0.3%
674
+0.10
0.3%
1224
+0.09
0.3%
Correlated Neurons
Index
P. Corr.
Cos Sim.
227
+0.11
0.06
113
+0.10
0.04
138
+0.09
0.06
Negative Logits
lele
-1.08
makro
-1.04
saar
-1.01
pama
-1.00
maksi
-1.00
teras
-0.99
akku
-0.99
seksi
-0.99
kafe
-0.98
bandung
-0.94
POSITIVE LOGITS
<bos>
1.42
expandindo
0.85
AutoScaleMode
0.71
SOUNDBITE
0.69
resourceCulture
0.68
Erreferentziak
0.66
betweenstory
0.61
PyExc
0.60
webElementXpaths
0.58
glLoadIdentity
0.58
Activations Density 0.354%