ResearchDatasetsPooled SharedFollowBenchmarksTask NameDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyTask NameDataset NameSOTA ResultTrendResultsVision-Language Question AnsweringPooled Shared (GQA, POPE, LLaVA-Wild, MMMU Pro, GMAI-MMBench, MME-Finance) (test)5.4Expected Calibration Error (ECE)22
Vision-Language Question AnsweringPooled Shared (GQA, POPE, LLaVA-Wild, MMMU Pro, GMAI-MMBench, MME-Finance) (test)5.4Expected Calibration Error (ECE)22