INDEX
Explanations
the definite article "the."
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
233
+0.13
0.7%
133
+0.12
0.6%
346
+0.11
0.6%
Correlated Neurons
Index
P. Corr.
Cos Sim.
10
+0.13
0.47
98
+0.12
0.21
228
+0.11
0.32
Negative Logits
"}](#
-1.67
ocations
-1.58
labelled
-1.56
arcin
-1.56
npmjs
-1.49
Errno
-1.49
reality
-1.46
undefined
-1.44
amaz
-1.41
aby
-1.37
POSITIVE LOGITS
º
2.85
ģ
2.67
ĸ´
2.62
Īĺ
2.56
Ĥ¬
2.55
Ļª
2.53
»
2.53
ª
2.49
Ļ
2.48
ļ
2.47
Activations Density 2.922%