INDEX
Explanations
information related to loading operations or tasks
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
1691
+0.11
0.4%
298
+0.10
0.4%
1573
+0.10
0.4%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1691
+0.11
0.03
1573
+0.10
0.02
976
+0.10
0.03
Negative Logits
Kinn
-0.45
drav
-0.45
Böh
-0.45
Tuscany
-0.44
Fran
-0.43
Francis
-0.42
Squal
-0.41
Stavanger
-0.41
Kristian
-0.41
pancre
-0.41
POSITIVE LOGITS
load
1.36
Load
1.30
loading
1.30
load
1.29
Loading
1.23
loads
1.21
Load
1.21
loading
1.20
loads
1.19
loaded
1.16
Activations Density 0.099%