INDEX
Explanations
phrases related to a specific dish, "quiche"
phrases related to clichés
New Auto-Interp
Negative Logits
elephants
-0.75
RED
-0.75
Relief
-0.65
org
-0.63
sorted
-0.63
plants
-0.61
monkeys
-0.61
cats
-0.60
relief
-0.60
SIGN
-0.60
POSITIVE LOGITS
iche
4.86
ichen
1.27
ICH
1.18
ich
1.16
inoa
1.05
ic
1.05
ikan
1.05
ique
1.01
icist
0.96
isle
0.95
Activations Density 0.014%