2024
Rethinking the Evaluation of Out-of-Distribution Detection: A Sorites Paradox
NeurIPS 2024poster
Most existing out-of-distribution (OOD) detection benchmarks classify samples with novel labels as the OOD data. However, some marginal OOD samples actually have close semantic contents to the in-distribution (ID) sample, which makes determining the OOD sample a Sorites Paradox. In this paper, we co…