INDEX
Explanations
references to clubs and organizations
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
156
+0.24
1.5%
507
+0.16
1.0%
362
+0.15
0.9%
Correlated Neurons
Index
P. Corr.
Cos Sim.
507
+0.24
0.02
362
+0.16
0.02
412
+0.15
0.02
Negative Logits
ĥ½
-2.51
§
-2.46
¯
-2.42
ĩ
-2.31
ĨĴ
-2.16
¤
-2.07
ī
-2.04
¸
-1.99
ĵ
-1.98
ĸ
-1.98
POSITIVE LOGITS
ycin
2.07
doms
2.06
aña
2.00
bing
1.94
soc
1.83
forum
1.82
Awards
1.80
house
1.74
smen
1.72
san
1.71
Activations Density 0.087%