INDEX
Explanations
references to air-related topics and objects
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
50
+0.20
1.3%
1296
+0.15
1.0%
1937
+0.10
0.7%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1296
+0.20
0.05
892
+0.15
0.05
1381
+0.10
0.03
Negative Logits
<bos>
-3.21
ⓧ
-0.70
#![
-0.68
struct
-0.67
lateinit
-0.67
CascadeType
-0.66
mobilize
-0.66
renovate
-0.65
relocate
-0.65
encourage
-0.65
POSITIVE LOGITS
véhic
1.39
soulign
1.37
Minang
1.31
bandung
1.28
applau
1.26
quoique
1.23
lele
1.22
maksi
1.22
déchir
1.20
Traité
1.20
Activations Density 0.123%