ResearchDatasetsharmful requests and jailbreak promptsFollowBenchmarksTask NameDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyTask NameDataset NameSOTA ResultTrendResultsAttack Success Rate20,000 harmful requests and 20,000 jailbreak prompts (test)80.2Attack Success Rate (ASR)18
Attack Success Rate20,000 harmful requests and 20,000 jailbreak prompts (test)80.2Attack Success Rate (ASR)18