ResearchBenchmarksImage Classification on unmet-promise (Split 1)Follow57.3Task PerformanceLens 7B53.24454.29755.3556.403Nov 6, 2025Evaluation ResultsMethodMethodLinksTask PerformanceLens 7Bvariant=debiasedvariant=debiased2025.1157.3MMD²2025.1157.3Test mean2025.1157.2MDM2025.1157Lens 7Bvariant=biasedvariant=biased2025.1156.7Lens 32Bvariant=debiasedvariant=debiased2025.1156.4Mauve2025.1156PAD2025.1155.9Lens 32Bvariant=biasedvariant=biased2025.1153.4