INDEX
Explanations
pronoun or name followed by speech verb
New Auto-Interp
Negative Logits
把你
0.78
}}/
0.73
ทำการ
0.71
uduk
0.71
만드는
0.69
获取
0.68
verlassen
0.67
askell
0.67
}}:
0.67
கூடாது
0.67
POSITIVE LOGITS
said
2.46
remarked
2.26
exclaimed
2.14
said
2.05
replied
2.02
stated
1.92
commented
1.91
says
1.89
explained
1.80
declared
1.76
Activations Density 0.058%