ResearchDatasetsHellaSwag, PIQA, SIQA, WinoGrandeHellaSwag, PIQA, SIQA, WinoGrandeFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsCommonsense ReasoningHellaSwag (HS), PIQA, SIQA, WinoGrande (Wino)78.8HellaSwag Accuracy13