ResearchDatasetsLINGUASAFEFollowBenchmarksTask NameDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyTask NameDataset NameSOTA ResultTrendResultsHarmlessness EvaluationLINGUASAFE (test)100Tie Rate38Helpfulness evaluationLINGUASAFE73.3Win Rate14