INDEX
Explanations
phrases or expressions related to intense or challenging experiences
New Auto-Interp
Negative Logits
emean
-0.17
okie
-0.15
ayo
-0.14
rss
-0.14
emente
-0.14
yms
-0.14
tees
-0.14
eyi
-0.14
reesome
-0.13
mtree
-0.13
POSITIVE LOGITS
ishly
0.31
ish
0.30
fire
0.30
acious
0.28
zap
0.26
raising
0.26
hole
0.26
hath
0.25
hound
0.24
rais
0.24
Activations Density 0.014%