INDEX
Explanations
words associated with technical processes and systems
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
365
+0.14
0.8%
227
+0.12
0.7%
271
+0.11
0.6%
Correlated Neurons
Index
P. Corr.
Cos Sim.
433
+0.14
0.06
292
+0.12
0.12
335
+0.11
0.12
Negative Logits
similarities
-1.72
¹
-1.48
trends
-1.48
liabilities
-1.38
plenty
-1.34
closely
-1.33
anomalies
-1.33
appearances
-1.31
indications
-1.30
crew
-1.29
POSITIVE LOGITS
-------------------------------------------------------
1.64
attention
1.64
áĢ¬
1.57
------------------------------------------------------
1.55
ãĤį
1.55
----------------------------------------------------
1.49
---------------------------------------------------
1.44
ainer
1.41
--------------------------------------------------
1.41
entre
1.39
Activations Density 4.996%