INDEX
Explanations
names of companies, products, and places
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
1842
+0.15
0.4%
1150
+0.15
0.4%
227
+0.14
0.4%
Correlated Neurons
Index
P. Corr.
Cos Sim.
227
+0.15
0.08
690
+0.15
0.03
981
+0.14
0.06
Negative Logits
utensi
-0.68
Paglinawan
-0.64
podjela
-0.63
MigrationBuilder
-0.61
بيها
-0.61
quei
-0.60
furg
-0.60
rilass
-0.59
HttpNotFound
-0.59
quese
-0.58
POSITIVE LOGITS
depic
1.43
impra
1.43
encomp
1.41
maneu
1.40
vainly
1.39
unspeak
1.38
quitted
1.36
intersper
1.36
gaily
1.34
tolerably
1.34
Activations Density 0.383%