ResearchTasksMulti-task Image-Text UnderstandingMulti-task Image-Text UnderstandingFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedVQAV2, GQA, NLVR2, COCO CaptionFT66.9VQAv2 Accuracy3Feb 26, 2026