INDEX
Explanations
mentions of the city of St. Louis
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
145
+0.14
0.8%
245
+0.13
0.7%
98
+0.11
0.6%
Correlated Neurons
Index
P. Corr.
Cos Sim.
463
+0.14
0.02
23
+0.13
0.02
456
+0.11
0.00
Negative Logits
ause
-2.17
rael
-1.96
aways
-1.60
away
-1.60
hered
-1.59
inia
-1.56
bows
-1.55
arily
-1.52
initis
-1.49
ulsions
-1.46
POSITIVE LOGITS
ĸ
5.63
Ń
5.43
¾
5.42
Ī
5.35
į
5.30
Ķ
5.26
Į
5.07
Ŀ
5.05
°
5.04
ĥ
5.04
Activations Density 0.903%