INDEX
Explanations
references to items or concepts characterized as "medium" in size or significance
New Auto-Interp
Negative Logits
ment
-0.18
anal
-0.16
zym
-0.15
zen
-0.15
mma
-0.14
iaux
-0.14
iger
-0.14
anter
-0.14
actory
-0.14
ides
-0.14
POSITIVE LOGITS
-sized
0.42
sized
0.41
Sized
0.31
-size
0.28
-large
0.23
size
0.23
ship
0.23
-term
0.23
-priced
0.23
-range
0.22
Activations Density 0.020%