INDEX
Explanations
phrases indicating trends or popular cultural shifts
New Auto-Interp
Negative Logits
EDIA
-0.15
antom
-0.15
LOPT
-0.15
acd
-0.14
enos
-0.14
eka
-0.14
anton
-0.14
teb
-0.14
allest
-0.14
ozem
-0.13
POSITIVE LOGITS
fashion
0.61
fashionable
0.54
-fashion
0.52
Fashion
0.51
trendy
0.48
cool
0.47
trend
0.44
Fashion
0.44
Cool
0.43
hip
0.43
Activations Density 0.157%