ResearchTasksDownstream task executionDownstream task executionFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedAgentSkillOSSKILLRAE84.59Score6May 12, 2026SkillsBenchSKILLRAE29.26Reward Mean (%)6May 12, 2026