INDEX
Explanations
locations mentioned in a news context
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
453
+0.20
0.6%
964
+0.20
0.6%
906
+0.12
0.4%
Correlated Neurons
Index
P. Corr.
Cos Sim.
453
+0.20
0.06
964
+0.20
0.04
1241
+0.12
0.04
Negative Logits
milf
-2.32
increa
-2.31
hairc
-2.31
ftu
-2.30
scrat
-2.24
affor
-2.23
fta
-2.23
disagre
-2.23
excru
-2.22
reluct
-2.21
POSITIVE LOGITS
***!
1.02
NSCoder
0.95
snippetHide
0.89
AnchorStyles
0.86
PerformLayout
0.86
EndContext
0.84
rawDesc
0.83
ContentAlignment
0.83
GraphicsUnit
0.82
كومونز
0.81
Activations Density 0.213%