INDEX
Explanations
phrases related to military vehicles or tools
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
492
+0.16
0.7%
812
+0.13
0.6%
1233
+0.13
0.5%
Correlated Neurons
Index
P. Corr.
Cos Sim.
492
+0.16
0.04
1233
+0.13
0.03
75
+0.13
0.02
Negative Logits
įsi
-0.52
laiko
-0.47
atve
-0.44
kira
-0.44
oudh
-0.43
luka
-0.43
saad
-0.42
teis
-0.42
caucasian
-0.41
maksi
-0.41
POSITIVE LOGITS
Pan
1.34
pan
1.32
Pan
1.31
PAN
1.30
pans
1.16
pan
1.15
Pans
1.13
PAN
1.06
panning
1.01
Panthers
0.94
Activations Density 0.079%