ConceptIndependent Fair Labels

Independent Fair Labels

A data assumption stating that the label of each sample is drawn independently from a uniform distribution, ensuring that no exploitable regularity exists in the data.

Timeline (last 90 days)

Oct 15

The counterexample is constructed using independent fair labels (each sampled label is equally likely to be positive or negative) and a permutation-equivariant learner.

Unverified50%

All Facts (1)

Unverified

The counterexample is constructed using independent fair labels (each sampled label is equally likely to be positive or negative) and a permutation-equivariant learner.

50%

Source Articles

Memorization Lower Bound Overturned: Posterior Frequencies Yield Exact CorrectionOct 15