Is the metric False Negative Variance the less the better?

Is False Negative Variance = False Negative/std? Is it means that high False Negative Variance represents a bad and stable performence? Please correct me if I was wrong.

Yes exactly, FN Variance = FN*(std over the whole cohort). Hence a high FN Variance indicates an unstable performance. The lower the metric the better, i.e. the lower the std of FN over the cohort the better.

but what if my FN is extremely high and stable, for example: FN = 100,100,100,100,100,99.9,100.1 and my FN varience would be extremely high, would that means a good performence? or my FN looks like : 0.1, 0.1, 0.1, 0.2, 0.1, 0.1, 0.1, the metric would be worse than the first one?

Thank you for bringing this to our attention. We have put further clarifications in the Evaluation page and opted for more meaningful names.