ResearchDatasetsAQUA-RAT, HumanEval, and GSM8KFollowBenchmarksTask NameDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyTask NameDataset NameSOTA ResultTrendResultsMulti-task Reasoning and EfficiencyAQUA-RAT, HumanEval, and GSM8K Aggregate60.9Average Accuracy8