ResearchDatasetsGPT-4 assisted benchmarkFollowBenchmarksTask NameDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyTask NameDataset NameSOTA ResultTrendResultsVisual Hallucination MitigationGPT-4 assisted benchmark46.7SHR15