INDEX
Explanations
people's names and specific events mentioned in news articles
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
184
+0.09
0.3%
1940
+0.08
0.2%
904
+0.08
0.2%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1471
+0.09
0.03
184
+0.08
0.01
981
+0.08
0.04
Negative Logits
Roskov
-0.72
WithIOException
-0.67
disambiguazione
-0.64
Walkover
-0.64
Autoritní
-0.64
Solución
-0.62
bolí
-0.61
noten
-0.61
Panamoan
-0.60
ddelweddau
-0.59
POSITIVE LOGITS
maneu
1.25
depic
1.13
shenan
1.10
unspeak
1.07
homeward
1.07
indestru
1.06
apprehen
1.05
encomp
1.05
unve
1.05
gaily
1.04
Activations Density 0.154%