INDEX
Explanations
code snippets related to data manipulation and calculations
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
764
+0.16
0.5%
964
+0.15
0.5%
876
+0.15
0.4%
Correlated Neurons
Index
P. Corr.
Cos Sim.
876
+0.16
-0.00
523
+0.15
0.02
906
+0.15
-0.00
Negative Logits
Singapur
-0.81
kosme
-0.81
optik
-0.78
kompres
-0.78
kooper
-0.75
silikon
-0.75
Ukraina
-0.73
akut
-0.72
Demok
-0.71
kanad
-0.69
POSITIVE LOGITS
nutella
0.86
polenta
0.69
ciao
0.69
purée
0.69
sappi
0.67
mascarpone
0.65
desideri
0.64
larged
0.63
giusti
0.63
parma
0.62
Activations Density 0.194%