INDEX
Explanations
mentions of countries or specific locations
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
227
+0.12
0.3%
1013
+0.08
0.2%
1862
+0.07
0.2%
Correlated Neurons
Index
P. Corr.
Cos Sim.
227
+0.12
0.07
284
+0.08
0.06
1013
+0.07
0.06
Negative Logits
reger
-0.71
logis
-0.67
erit
-0.62
tenden
-0.61
ordina
-0.60
gesta
-0.59
tages
-0.58
vola
-0.57
tuc
-0.57
hek
-0.57
POSITIVE LOGITS
shortly
0.67
posób
0.66
pymongo
0.63
pymysql
0.60
early
0.59
last
0.59
Shortly
0.59
late
0.58
earlier
0.57
בשנת
0.57
Activations Density 0.472%