INDEX
Explanations
sentences related to military or violent events
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
381
+0.16
0.5%
1741
+0.12
0.4%
776
+0.12
0.4%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1919
+0.16
0.07
331
+0.12
0.05
381
+0.12
0.03
Negative Logits
sentito
-0.89
tuong
-0.79
cristi
-0.79
monaster
-0.78
affez
-0.75
sacerd
-0.74
mosso
-0.74
toscana
-0.74
nguyen
-0.74
venuto
-0.74
POSITIVE LOGITS
infatti
0.73
totiž
0.72
assailed
0.70
impelled
0.69
Bardzo
0.68
Namely
0.68
endeavouring
0.67
bowiem
0.67
endeavored
0.64
strove
0.64
Activations Density 0.900%