ResearchDatasetsComVEComVEFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsCommonsense Validation and ExplanationComVE (test)98.5Accuracy13Natural Language Explanation GenerationComVE70Human Evaluation Score7Generative Commonsense ReasoningComVE78.9Win-Tie Score3Natural Language Explanation GenerationComVE few-shot 60-shot68.03Accuracy3Commonsense Validation and ExplanationComVE0.88Performance (F+B -> P+B)2