ResearchTasksMulti-task Performance Change AnalysisMulti-task Performance Change AnalysisFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedGSM8K, MATH-500, MMLU, CNN/DMTop-k4.01Mean Delta5May 28, 2026