# LLM Jailbreaking on HarmBench text (test N = 320) Source: [LLM Jailbreaking on HarmBench text (test N = 320)]() Collection index: [llms.txt]() Data updated: 2026-04-29T07:21:43.939Z Evaluation of model susceptibility to jailbreaking attacks using the specific 320-sample subset of the HarmBench test set, measured by Attack Success Rate variants (ASR-M and ASR-J). 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 | | Dataset (including any stated split) | HarmBench text (test N = 320) | | Metric | ASR-M | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 16 of 16 | | 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 - [ASR-M](): Higher is better; unit: PERCENT. - [ASR-J](): Higher is better; unit: PERCENT. ## Recorded results | Method | ASR-M | Result date | Report paper | | --- | --- | --- | --- | | PEO | 93.75 | Not recorded | 69f1907555a3c32af89fb9a8 | | SPT | 92.19 | Not recorded | 69f1907555a3c32af89fb9a8 | | PEO | 90.94 | Not recorded | 69f1907555a3c32af89fb9a8 | | nanoGCG | 86.25 | Not recorded | 69f1907555a3c32af89fb9a8 | | PEO | 82.81 | Not recorded | 69f1907555a3c32af89fb9a8 | | SPT | 80 | Not recorded | 69f1907555a3c32af89fb9a8 | | PEO | 70.31 | Not recorded | 69f1907555a3c32af89fb9a8 | | BEAST | 69.38 | Not recorded | 69f1907555a3c32af89fb9a8 | | BEAST | 68.44 | Not recorded | 69f1907555a3c32af89fb9a8 | | SPT | 66.56 | Not recorded | 69f1907555a3c32af89fb9a8 | | nanoGCG | 64.69 | Not recorded | 69f1907555a3c32af89fb9a8 | | nanoGCG | 58.44 | Not recorded | 69f1907555a3c32af89fb9a8 | | BEAST | 53.44 | Not recorded | 69f1907555a3c32af89fb9a8 | | SPT | 47.19 | Not recorded | 69f1907555a3c32af89fb9a8 | | nanoGCG | 40 | Not recorded | 69f1907555a3c32af89fb9a8 | | BEAST | 39.06 | Not recorded | 69f1907555a3c32af89fb9a8 | ### PEO ASR-M: 93.75. Result ID: 69f1b0d047f34ce924d7bf84. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama-3.2-3B-Instruct"}. This record does not assert independent reproduction. ### SPT ASR-M: 92.19. Result ID: 69f1b0d047f34ce924d7bf86. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Llama-3.2-3B-Instruct"}. This record does not assert independent reproduction. ### PEO ASR-M: 90.94. Result ID: 69f1b0d047f34ce924d7bf78. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Vicuna-7B v1.3"}. This record does not assert independent reproduction. ### nanoGCG ASR-M: 86.25. Result ID: 69f1b0d047f34ce924d7bf79. [Reporting paper]() Recorded evaluation configuration: {"Model":"Vicuna-7B v1.3"}. This record does not assert independent reproduction. ### PEO ASR-M: 82.81. Result ID: 69f1b0d047f34ce924d7bf7c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama-2-7B-Chat"}. This record does not assert independent reproduction. ### SPT ASR-M: 80. Result ID: 69f1b0d047f34ce924d7bf7a. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Vicuna-7B v1.3"}. This record does not assert independent reproduction. ### PEO ASR-M: 70.31. Result ID: 69f1b0d047f34ce924d7bf80. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen3-4B-Instruct"}. This record does not assert independent reproduction. ### BEAST ASR-M: 69.38. Result ID: 69f1b0d047f34ce924d7bf7b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Vicuna-7B v1.3"}. This record does not assert independent reproduction. ### BEAST ASR-M: 68.44. Result ID: 69f1b0d047f34ce924d7bf87. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama-3.2-3B-Instruct"}. This record does not assert independent reproduction. ### SPT ASR-M: 66.56. Result ID: 69f1b0d047f34ce924d7bf7e. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Llama-2-7B-Chat"}. This record does not assert independent reproduction. ### nanoGCG ASR-M: 64.69. Result ID: 69f1b0d047f34ce924d7bf85. [Reporting paper]() Recorded evaluation configuration: {"Model":"Llama-3.2-3B-Instruct"}. This record does not assert independent reproduction. ### nanoGCG ASR-M: 58.44. Result ID: 69f1b0d047f34ce924d7bf7d. [Reporting paper]() Recorded evaluation configuration: {"Model":"Llama-2-7B-Chat"}. This record does not assert independent reproduction. ### BEAST ASR-M: 53.44. Result ID: 69f1b0d047f34ce924d7bf83. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen3-4B-Instruct"}. This record does not assert independent reproduction. ### SPT ASR-M: 47.19. Result ID: 69f1b0d047f34ce924d7bf82. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Qwen3-4B-Instruct"}. This record does not assert independent reproduction. ### nanoGCG ASR-M: 40. Result ID: 69f1b0d047f34ce924d7bf81. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen3-4B-Instruct"}. This record does not assert independent reproduction. ### BEAST ASR-M: 39.06. Result ID: 69f1b0d047f34ce924d7bf7f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama-2-7B-Chat"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.