INDEX
Explanations
references to technology-related products and services
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
23
+0.19
1.1%
258
+0.18
1.0%
321
+0.15
0.9%
Correlated Neurons
Index
P. Corr.
Cos Sim.
258
+0.19
0.17
321
+0.18
0.11
23
+0.15
0.14
Negative Logits
"}](#
-2.00
himself
-1.87
quarters
-1.61
unpublished
-1.60
ius
-1.60
ural
-1.56
physicist
-1.53
ème
-1.51
"].
-1.50
bench
-1.50
POSITIVE LOGITS
č↵č↵
1.76
Common
1.68
#
1.68
FORE
1.61
[âĢ¦]
1.55
Their
1.54
COM
1.53
respectively
1.53
Additionally
1.53
####
1.52
Activations Density 1.136%