INDEX
Explanations
e-commerce related deals or promotions
New Auto-Interp
Neuron Alignment
Index
Value
% of L₁
50
+0.10
0.5%
991
+0.05
0.2%
1622
+0.05
0.2%
Correlated Neurons
Index
P. Corr.
Cos Sim.
1234
+0.10
0.04
163
+0.05
0.04
1811
+0.05
0.04
Negative Logits
<bos>
-1.70
public
-0.79
,
-0.68
//
-0.68
-0.67
.
-0.66
-0.66
int
-0.66
var
-0.66
endif
-0.65
POSITIVE LOGITS
maneu
2.21
affor
2.11
increa
2.01
accla
1.99
impra
1.92
disagre
1.89
scrat
1.84
excru
1.83
inev
1.82
strick
1.80
Activations Density 0.093%