ResearchTasksBlack-box LLM JailbreakingFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedHarmBench (test)TrailBlazer98.74ASR (%)8Feb 26, 2026AdvBench GPT-oss-20B target (test)TrailBlazer85.3ASR6Feb 26, 2026AdvBench Qwen 3-14B target (test)PAIR51.92ASR6Feb 26, 2026AdvBench Gemma 3-12B target (test)FlipAttack100ASR6Feb 26, 2026AdvBench Llama 3.2-11B target (test)TrailBlazer95.51ASR6Feb 26, 2026