ResearchDatasetsARC, BoolQ, HellaSwagFollowBenchmarksTask NameDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyTask NameDataset NameSOTA ResultTrendResultsZero-shot Task EvaluationARC-C, ARC-E, BoolQ, and HellaSwag69.48Accuracy28