INDEX
Explanations
phrases that emphasize comparison and contrast
New Auto-Interp
Negative Logits
croft
-0.17
oyo
-0.16
673
-0.16
Unc
-0.15
osp
-0.14
Xu
-0.14
ÄĻ
-0.14
andler
-0.14
Socorro
-0.14
979
-0.14
POSITIVE LOGITS
/topics
0.16
Äįet
0.16
.Shape
0.15
ÏģιÏĥ
0.15
reator
0.15
ollision
0.14
ibri
0.14
ัà¸ģà¸ģ
0.14
abound
0.14
.dtp
0.14
Activations Density 0.023%