ResearchTasksDiscriminative Hallucination EvaluationFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedAMBER (test)RC-DPO86.8Accuracy42Jul 7, 2026AMBER-dUE-DPO89.5F1 Score23Jun 1, 2026AMBERMESA84.3Accuracy18Apr 21, 2026AMBER DiscriminativeQwen2.5-VL 3B + NoisyGRPO90.3F1 Score9Apr 8, 2026