INDEX
Explanations
adjectives related to sophistication or elegance
occurrences of the word "fancy" and variations related to it
New Auto-Interp
Negative Logits
etts
-0.78
scl
-0.75
essee
-0.74
sen
-0.72
upon
-0.71
trust
-0.71
aston
-0.71
cano
-0.71
released
-0.70
arf
-0.70
POSITIVE LOGITS
fancy
1.19
pants
1.01
tail
0.81
dress
0.79
nifty
0.74
rous
0.72
fanc
0.68
notions
0.68
nap
0.67
pants
0.66
Activations Density 0.028%