# LLM Jailbreaking Attack on AdvBench (100 samples) Source: [LLM Jailbreaking Attack on AdvBench (100 samples)]() Collection index: [llms.txt]() Data updated: 2026-05-12T10:11:07.834Z Evaluation of LLM susceptibility to jailbreaking attacks using a 100-sample subset of the AdvBench dataset, reporting Attack Success Rate. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | LLM Jailbreaking Attack | | Dataset (including any stated split) | AdvBench (100 samples) | | Metric | Attack Success Rate | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 12 of 12 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Attack Success Rate](): Higher is better; unit: PERCENT. ## Recorded results | Method | Attack Success Rate | Result date | Report paper | | --- | --- | --- | --- | | Semantic Representation Attack | 100 | Not recorded | 6a02b409dbc836b87d82462b | | Semantic Representation Attack | 100 | Not recorded | 6a02b409dbc836b87d82462b | | Semantic Representation Attack | 96 | Not recorded | 6a02b409dbc836b87d82462b | | Semantic Representation Attack | 92 | Not recorded | 6a02b409dbc836b87d82462b | | AutoDAN | 90 | Not recorded | 6a02b409dbc836b87d82462b | | AutoDAN | 78 | Not recorded | 6a02b409dbc836b87d82462b | | PAIR | 76 | Not recorded | 6a02b409dbc836b87d82462b | | AutoDAN | 74 | Not recorded | 6a02b409dbc836b87d82462b | | PAIR | 62 | Not recorded | 6a02b409dbc836b87d82462b | | AutoDAN | 56 | Not recorded | 6a02b409dbc836b87d82462b | | PAIR | 52 | Not recorded | 6a02b409dbc836b87d82462b | | PAIR | 48 | Not recorded | 6a02b409dbc836b87d82462b | ### Semantic Representation Attack Attack Success Rate: 100. Result ID: 6a02fc8c47f34ce924e64e83. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Defense Type":"SmoothLLM-Swap"}. This record does not assert independent reproduction. ### Semantic Representation Attack Attack Success Rate: 100. Result ID: 6a02fc8c47f34ce924e64e85. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Defense Type":"SmoothLLM-Patch"}. This record does not assert independent reproduction. ### Semantic Representation Attack Attack Success Rate: 96. Result ID: 6a02fc8c47f34ce924e64e84. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Defense Type":"SmoothLLM-Insert"}. This record does not assert independent reproduction. ### Semantic Representation Attack Attack Success Rate: 92. Result ID: 6a02fc8c47f34ce924e64e82. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Defense Type":"Defenseless"}. This record does not assert independent reproduction. ### AutoDAN Attack Success Rate: 90. Result ID: 6a02fc8c47f34ce924e64e7e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Defense Type":"Defenseless"}. This record does not assert independent reproduction. ### AutoDAN Attack Success Rate: 78. Result ID: 6a02fc8c47f34ce924e64e80. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Defense Type":"SmoothLLM-Insert"}. This record does not assert independent reproduction. ### PAIR Attack Success Rate: 76. Result ID: 6a02fc8c47f34ce924e64e7a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Defense Type":"Defenseless"}. This record does not assert independent reproduction. ### AutoDAN Attack Success Rate: 74. Result ID: 6a02fc8c47f34ce924e64e81. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Defense Type":"SmoothLLM-Patch"}. This record does not assert independent reproduction. ### PAIR Attack Success Rate: 62. Result ID: 6a02fc8c47f34ce924e64e7c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Defense Type":"SmoothLLM-Insert"}. This record does not assert independent reproduction. ### AutoDAN Attack Success Rate: 56. Result ID: 6a02fc8c47f34ce924e64e7f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Defense Type":"SmoothLLM-Swap"}. This record does not assert independent reproduction. ### PAIR Attack Success Rate: 52. Result ID: 6a02fc8c47f34ce924e64e7d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Defense Type":"SmoothLLM-Patch"}. This record does not assert independent reproduction. ### PAIR Attack Success Rate: 48. Result ID: 6a02fc8c47f34ce924e64e7b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Defense Type":"SmoothLLM-Swap"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.