ResearchDatasetsGSM8K, CommonSense, BoolQ, ARC, and HellaSwagFollowBenchmarksTask NameDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyTask NameDataset NameSOTA ResultTrendResultsReasoning and MathGSM8K, CommonSense, BoolQ, ARC Challenge, and HellaSwag87.8Average Accuracy9