INDEX
Explanations
LaTeX code referencing equation labels with scaling
New Auto-Interp
Negative Logits
↵↵
-1.70
↵
-1.03
<bos>
-0.94
<eos>
-0.91
-0.89
↵↵↵
-0.87
The
-0.83
’
-0.82
-0.81
“
-0.77
POSITIVE LOGITS
BnF
0.80
end
0.77
CWE
0.75
Himo
0.71
Coder
0.69
NSP
0.67
Paglinawan
0.66
NDL
0.65
ENO
0.64
Minion
0.64
Activations Density 4.090%