ResearchTasksTATAFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedCurated Prompt Set (OOD)VideoPhy0.605RM Accuracy8Feb 26, 2026Internal ID (train)SoliReward0.79RM ACC8Feb 26, 2026