INDEX
Explanations
words related to physical body parts, especially the belly
references to physical attributes and actions related to the abdomen
New Auto-Interp
Negative Logits
inational
-0.77
inatory
-0.76
inates
-0.76
icity
-0.73
itars
-0.73
inations
-0.72
å§«
-0.71
atography
-0.71
ophical
-0.68
URA
-0.68
POSITIVE LOGITS
belly
1.20
button
0.93
belt
0.87
bum
0.82
GROUND
0.81
nect
0.79
seat
0.79
loader
0.77
fish
0.73
whale
0.72
Activations Density 0.012%