INDEX
Explanations
terms related to meals or dining occasions
mentions of feasting or related festivities
New Auto-Interp
Negative Logits
ple
-0.67
nuclear
-0.66
lost
-0.63
por
-0.63
abusive
-0.63
sold
-0.62
personal
-0.61
pe
-0.61
peer
-0.60
cel
-0.60
POSITIVE LOGITS
feast
1.27
Feast
1.22
oleon
0.92
efully
0.89
ctuary
0.83
attRot
0.83
archy
0.83
rite
0.82
supper
0.80
ivals
0.79
Activations Density 0.010%