INDEX
Explanations
information and instructions related to software or technology
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
381
+0.16
0.5%
1510
+0.13
0.4%
1919
+0.12
0.4%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1919
+0.16
0.09
1415
+0.13
0.06
30
+0.12
0.07
Negative Logits
gubern
-0.72
autunno
-0.71
Huhu
-0.71
LXXX
-0.71
poliester
-0.71
affari
-0.70
Venise
-0.69
raffredd
-0.69
Poitiers
-0.69
Chinois
-0.68
POSITIVE LOGITS
needn
0.82
You
0.76
mustn
0.75
You
0.74
yourself
0.73
you
0.71
can
0.71
might
0.70
should
0.70
may
0.69
Activations Density 0.225%