INDEX
Explanations
references to counterfeit and luxury fashion items
New Auto-Interp
Negative Logits
propOrder
-0.82
-0.81
surla
-0.73
featureID
-0.72
صوتيه
-0.69
pinulongan
-0.68
للاسماء
-0.66
تقاوى
-0.65
RenderAtEndOf
-0.64
GenerationType
-0.64
POSITIVE LOGITS
uitton
0.65
vuitton
0.62
replica
0.57
mole
0.54
Gucci
0.52
Vuitton
0.51
replicas
0.51
fakes
0.49
goods
0.48
bootleg
0.47
Activations Density 0.086%