ResearchTasksHarmful QueryFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedSORRYPeCAN12.79ASR (%)20Feb 26, 2026JailbreakBNSPO0.67ASR20Feb 26, 2026HarmBNSPO0.18ASR (%)20Feb 26, 2026PKU-SafeNSPO0.52ASR20Feb 26, 2026