INDEX
Explanations
words related to political and humanitarian crises, especially in a specific region
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
604
+0.14
0.4%
468
+0.12
0.4%
1394
+0.11
0.3%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1394
+0.14
0.07
939
+0.12
0.08
713
+0.11
0.05
Negative Logits
<bos>
-1.39
logitech
-1.00
funko
-0.86
unsplash
-0.85
swarovski
-0.83
airpods
-0.80
lancia
-0.80
vété
-0.78
vieilles
-0.77
Chinois
-0.75
POSITIVE LOGITS
ongoing
0.53
conflict
0.52
humanitarian
0.51
новниш
0.47
refugees
0.45
population
0.44
war
0.43
zeera
0.43
سكانية
0.43
civilians
0.43
Activations Density 0.649%