INDEX
Explanations
words related to scents or aromas
New Auto-Interp
Negative Logits
ant
-0.22
оÑı
-0.20
ignet
-0.19
op
-0.17
by
-0.17
am
-0.16
andal
-0.16
ammer
-0.16
stacle
-0.16
selling
-0.16
POSITIVE LOGITS
sc
0.27
ented
0.21
ree
0.20
otch
0.19
otland
0.19
issor
0.18
ents
0.18
arily
0.17
uba
0.17
ount
0.17
Activations Density 0.005%