INDEX
Negative Logits
ii
1.24
II
1.22
II
1.21
ii
1.13
III
1.01
iii
0.95
iii
0.92
Ii
0.92
Ⅱ
0.91
III
0.89
POSITIVE LOGITS
iv
1.51
IV
1.33
Iv
1.29
Iv
1.23
iv
1.19
IV
1.07
ivg
1.03
riv
0.97
ivé
0.96
ive
0.96
Activations Density 0.010%
ii
II
II
ii
III
iii
iii
Ii
Ⅱ
III
iv
IV
Iv
Iv
iv
IV
ivg
riv
ivé
ive