INDEX
Explanations
culinary techniques and preparation methods
New Auto-Interp
Negative Logits
ekk
-0.18
arend
-0.14
agreeing
-0.14
견
-0.14
Providing
-0.14
advocating
-0.13
kaar
-0.13
leading
-0.13
ekl
-0.13
dando
-0.13
POSITIVE LOGITS
cutting
0.20
rins
0.20
squir
0.20
sco
0.19
drying
0.19
clipping
0.19
kne
0.19
flatt
0.18
drilling
0.18
washing
0.17
Activations Density 0.216%