ResearchTaskscommonsense inferenceFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedHellaSwagN-3-Super 120B-A12B-Base88.97Accuracy171Jun 18, 2026HellaSwag (test)FP1684.89Accuracy4Feb 26, 2026