ResearchDatasetsGSM8K, MATH, MMLU, CNN/DMGSM8K, MATH, MMLU, CNN/DMFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsMulti-task Performance Change AnalysisGSM8K, MATH-500, MMLU, CNN/DM4.01Mean Delta5