ResearchTasksHarmfulness AssessmentHarmfulness AssessmentFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedBi-directional Anchoring evaluation setVaccine-SFT28.5Harmful Score6Feb 26, 2026AJailBenchSARSteer18ASR3May 8, 2026