INDEX
Explanations
descriptions related to specific events or topics with numerical references
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
1896
+0.18
0.6%
1967
+0.16
0.5%
453
+0.13
0.4%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1896
+0.18
0.03
1034
+0.16
0.04
1967
+0.13
0.03
Negative Logits
milf
-1.35
snoopy
-1.34
stockholm
-1.30
strick
-1.28
shenan
-1.28
depic
-1.27
jurassic
-1.26
beaute
-1.26
madonna
-1.25
unve
-1.24
POSITIVE LOGITS
dozen
0.78
couple
0.77
dozen
0.76
few
0.75
million
0.69
lot
0.68
billion
0.64
handful
0.64
decade
0.64
Enllaços
0.62
Activations Density 0.212%