INDEX
Explanations
mentions of specific selections or choices, such as those related to music labels, TV shows, or tournament participants
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
1758
+0.16
0.6%
521
+0.14
0.5%
849
+0.11
0.4%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1758
+0.16
0.03
521
+0.14
0.02
1479
+0.11
0.02
Negative Logits
***!
-0.54
TextFormField
-0.50
cambrian
-0.44
Campionato
-0.43
médaille
-0.43
Britton
-0.42
tamia
-0.42
InstrumentedTest
-0.41
Lordship
-0.41
LookAnd
-0.41
POSITIVE LOGITS
selects
1.13
Selection
1.12
selection
1.11
selections
1.10
select
1.08
Select
1.04
selecting
1.03
Selecting
1.01
Selections
1.01
selected
0.98
Activations Density 0.068%