ResearchTasksImage SelectionFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedWinogroundSELFEVAL14Image Score7Feb 26, 2026Interleaved Image-Text Composition Human Preference StudyHuman1Score4Feb 27, 2026