INDEX
Negative Logits
hence
0.96
instead
0.94
Hence
0.93
ibid
0.90
dunque
0.86
Ibid
0.83
Therefore
0.83
Thus
0.82
Ibid
0.82
Hence
0.82
POSITIVE LOGITS
doesn
1.25
surprisingly
1.15
adapts
1.14
plenty
1.13
laughs
1.12
undeniably
1.12
Doesn
1.11
reminds
1.11
luôn
1.10
surprisingly
1.09
Activations Density 0.587%