INDEX
Explanations
relational phrases and structures in sentences
New Auto-Interp
Negative Logits
ONO
-0.17
YG
-0.16
\grid
-0.16
inia
-0.15
/REC
-0.14
ÅĻÃŃd
-0.14
outu
-0.14
ewise
-0.14
Å¡tÃŃ
-0.14
immel
-0.14
POSITIVE LOGITS
three
0.54
two
0.51
four
0.41
three
0.40
several
0.38
five
0.36
two
0.35
ä¸ī个
0.35
两个
0.34
six
0.33
Activations Density 0.272%