INDEX
Explanations
specific phrases related to meal types and dining occasions
New Auto-Interp
Negative Logits
roke
-0.15
Um
-0.15
dan
-0.14
tune
-0.14
Personen
-0.14
yne
-0.14
dash
-0.14
Maj
-0.13
Han
-0.13
wakes
-0.13
POSITIVE LOGITS
orny
0.16
adiens
0.15
RelativeTo
0.15
ordion
0.15
apus
0.15
ischer
0.14
Cath
0.14
bfd
0.14
rique
0.14
adia
0.14
Activations Density 0.170%