Acknowledges Derivation but Questions Overly Strong Conclusions

KordingLab · x · 2026-07-15

This reply acknowledges the mathematical derivation as 'valid and interesting', but expresses reservation: the statement on page 4 of the paper that 'which metric you choose doesn't really matter' is a stronger claim than the formal result. The author believes the existing formalization does not actually support such a strong conclusion.

Related event: Debate Sparks Over Metric Dependence in AI Research(4 posts)→

Original post →

More from Research

Research channel →