ResearchTasksCommonsense Sentence CompletionFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedHellaSwag 2K (test)G-DAUG-Combo76.8Accuracy6Feb 26, 2026CODAH (test)G-DAUG-Influence84.3Accuracy6Feb 26, 2026HellaSwag full (test val)GPT2-201M26.7Accuracy3May 21, 2026