ResearchDatasetsAIME, HMMT, HE, MBPPFollowBenchmarksTask NameDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyTask NameDataset NameSOTA ResultTrendResultsMulti-task Model EvaluationAIME24, AIME25, HMMT25, HE+, MBPP+ Average58.4Average Accuracy16