INDEX
Explanations
articles and determiners in the text
New Auto-Interp
Negative Logits
ForObject
-0.15
ixed
-0.14
yne
-0.14
Antar
-0.14
gig
-0.14
tel
-0.14
emmel
-0.14
BET
-0.13
of
-0.13
way
-0.13
POSITIVE LOGITS
/ws
0.15
SSIP
0.15
епÑĤи
0.15
ypress
0.14
uhl
0.14
дав
0.14
atoi
0.14
readcr
0.14
ooke
0.14
patches
0.14
Activations Density 0.072%