INDEX
Explanations
sweet food items, particularly those made with chocolate
references to chocolate and related desserts
New Auto-Interp
Negative Logits
plur
-0.75
WARD
-0.72
umbnail
-0.70
href
-0.69
kus
-0.68
ership
-0.67
upload
-0.67
Filename
-0.67
REPORT
-0.67
NER
-0.64
POSITIVE LOGITS
pudding
1.01
anut
0.92
cake
0.89
chip
0.86
cup
0.86
cane
0.85
flake
0.85
coated
0.84
syrup
0.84
flavored
0.83
Activations Density 0.025%