INDEX
Explanations
snack-related words or phrases
references to snack foods
New Auto-Interp
Negative Logits
ocally
-0.80
negatives
-0.72
ocal
-0.68
ttle
-0.67
Æ
-0.66
idency
-0.64
nea
-0.64
plur
-0.63
otype
-0.63
20439
-0.61
POSITIVE LOGITS
foods
0.88
Food
0.86
snack
0.85
cereal
0.84
eaten
0.84
snacks
0.83
Eating
0.83
food
0.82
washer
0.80
food
0.76
Activations Density 0.041%