ResearchDatasetsJBB, AdvBench, and HarmBenchFollowBenchmarksTask NameDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyTask NameDataset NameSOTA ResultTrendResultsAudio JailbreakJBB, AdvBench, and HarmBench Combined98.16ASR (Qwen 2-Audio)10