INDEX
Explanations
words related to competition or competitive contexts
New Auto-Interp
Negative Logits
pr
-0.17
latex
-0.16
psilon
-0.16
pha
-0.16
labs
-0.15
prs
-0.15
inals
-0.15
osen
-0.15
atics
-0.14
ongoose
-0.14
POSITIVE LOGITS
comp
0.27
Comp
0.25
ensation
0.23
.comp
0.23
-comp
0.23
(comp
0.23
aign
0.20
ounding
0.20
licated
0.20
.Comp
0.20
Activations Density 0.010%