ResearchTasksVQA HallucinationFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedMMHalGPT-4o3.87Score21Apr 6, 2026POPE MS COCO v1 (test)Robust mPLUG-Owl-7B86Random Accuracy11Feb 26, 2026