# Jailbreaking on JBB Behaviors (ASR, StR) Source: [Jailbreaking on JBB Behaviors (ASR, StR)]() Collection index: [llms.txt]() Data updated: 2026-04-03T07:05:20.332Z Evaluation of LLM performance against jailbreak prompts on the JBB Behaviors dataset, measured using Attack Success Rate (ASR) and Success Transfer Rate (StR). 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 | Jailbreaking | | Dataset (including any stated split) | JBB Behaviors | | Metric | ASR | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 35 of 35 | | 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](): Higher is better; unit: PERCENT. - [StR](): Higher is better; unit: PERCENT. ## Recorded results | Method | ASR | Result date | Report paper | | --- | --- | --- | --- | | AMIS | 100 | Not recorded | 690988963ef7ccd74e5e5efc | | AMIS | 100 | Not recorded | 690988963ef7ccd74e5e5efc | | PAP | 97 | Not recorded | 690988963ef7ccd74e5e5efc | | AMIS | 97 | Not recorded | 690988963ef7ccd74e5e5efc | | PAIR | 91 | Not recorded | 690988963ef7ccd74e5e5efc | | TAP | 91 | Not recorded | 690988963ef7ccd74e5e5efc | | SeqAR | 89 | Not recorded | 690988963ef7ccd74e5e5efc | | AMIS | 88 | Not recorded | 690988963ef7ccd74e5e5efc | | AutoDAN-Turbo | 85 | Not recorded | 690988963ef7ccd74e5e5efc | | PAP | 84 | Not recorded | 690988963ef7ccd74e5e5efc | | PAIR | 83 | Not recorded | 690988963ef7ccd74e5e5efc | | TAP | 80 | Not recorded | 690988963ef7ccd74e5e5efc | | AMIS | 78 | Not recorded | 690988963ef7ccd74e5e5efc | | PAIR | 77 | Not recorded | 690988963ef7ccd74e5e5efc | | TAP | 72 | Not recorded | 690988963ef7ccd74e5e5efc | | PAP | 69 | Not recorded | 690988963ef7ccd74e5e5efc | | PAP | 67 | Not recorded | 690988963ef7ccd74e5e5efc | | PAIR | 61 | Not recorded | 690988963ef7ccd74e5e5efc | | AutoDAN-Turbo | 60 | Not recorded | 690988963ef7ccd74e5e5efc | | TAP | 53 | Not recorded | 690988963ef7ccd74e5e5efc | | AutoDAN-Turbo | 45 | Not recorded | 690988963ef7ccd74e5e5efc | | Vanilla | 41 | Not recorded | 690988963ef7ccd74e5e5efc | | TAP | 37 | Not recorded | 690988963ef7ccd74e5e5efc | | AutoDAN-Turbo | 33 | Not recorded | 690988963ef7ccd74e5e5efc | | AutoDAN-Turbo | 31 | Not recorded | 690988963ef7ccd74e5e5efc | | PAIR | 29 | Not recorded | 690988963ef7ccd74e5e5efc | | PAP | 20 | Not recorded | 690988963ef7ccd74e5e5efc | | SeqAR | 16 | Not recorded | 690988963ef7ccd74e5e5efc | | SeqAR | 9 | Not recorded | 690988963ef7ccd74e5e5efc | | Vanilla | 3 | Not recorded | 690988963ef7ccd74e5e5efc | | Vanilla | 3 | Not recorded | 690988963ef7ccd74e5e5efc | | Vanilla | 2 | Not recorded | 690988963ef7ccd74e5e5efc | | Vanilla | 1 | Not recorded | 690988963ef7ccd74e5e5efc | | SeqAR | 0 | Not recorded | 690988963ef7ccd74e5e5efc | | SeqAR | 0 | Not recorded | 690988963ef7ccd74e5e5efc | ### AMIS ASR: 100. Result ID: 69cf66a047f34ce924b8bc8a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"Llama-3.1-8B"}. This record does not assert independent reproduction. ### AMIS ASR: 100. Result ID: 69cf66a047f34ce924b8bc8b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"GPT-4o-mini"}. This record does not assert independent reproduction. ### PAP ASR: 97. Result ID: 69cf66a047f34ce924b8bc7b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"Llama-3.1-8B"}. This record does not assert independent reproduction. ### AMIS ASR: 97. Result ID: 69cf66a047f34ce924b8bc8c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"GPT-4o"}. This record does not assert independent reproduction. ### PAIR ASR: 91. Result ID: 69cf66a047f34ce924b8bc71. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"Llama-3.1-8B"}. This record does not assert independent reproduction. ### TAP ASR: 91. Result ID: 69cf66a047f34ce924b8bc76. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"Llama-3.1-8B"}. This record does not assert independent reproduction. ### SeqAR ASR: 89. Result ID: 69cf66a047f34ce924b8bc80. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"Llama-3.1-8B"}. This record does not assert independent reproduction. ### AMIS ASR: 88. Result ID: 69cf66a047f34ce924b8bc8e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"Sonnet-4"}. This record does not assert independent reproduction. ### AutoDAN-Turbo ASR: 85. Result ID: 69cf66a047f34ce924b8bc85. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"Llama-3.1-8B"}. This record does not assert independent reproduction. ### PAP ASR: 84. Result ID: 69cf66a047f34ce924b8bc7c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"GPT-4o-mini"}. This record does not assert independent reproduction. ### PAIR ASR: 83. Result ID: 69cf66a047f34ce924b8bc72. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"GPT-4o-mini"}. This record does not assert independent reproduction. ### TAP ASR: 80. Result ID: 69cf66a047f34ce924b8bc77. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"GPT-4o-mini"}. This record does not assert independent reproduction. ### AMIS ASR: 78. Result ID: 69cf66a047f34ce924b8bc8d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"Haiku-3.5"}. This record does not assert independent reproduction. ### PAIR ASR: 77. Result ID: 69cf66a047f34ce924b8bc73. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"GPT-4o"}. This record does not assert independent reproduction. ### TAP ASR: 72. Result ID: 69cf66a047f34ce924b8bc78. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"GPT-4o"}. This record does not assert independent reproduction. ### PAP ASR: 69. Result ID: 69cf66a047f34ce924b8bc7d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"GPT-4o"}. This record does not assert independent reproduction. ### PAP ASR: 67. Result ID: 69cf66a047f34ce924b8bc7e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"Haiku-3.5"}. This record does not assert independent reproduction. ### PAIR ASR: 61. Result ID: 69cf66a047f34ce924b8bc74. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"Haiku-3.5"}. This record does not assert independent reproduction. ### AutoDAN-Turbo ASR: 60. Result ID: 69cf66a047f34ce924b8bc86. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"GPT-4o-mini"}. This record does not assert independent reproduction. ### TAP ASR: 53. Result ID: 69cf66a047f34ce924b8bc79. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"Haiku-3.5"}. This record does not assert independent reproduction. ### AutoDAN-Turbo ASR: 45. Result ID: 69cf66a047f34ce924b8bc87. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"GPT-4o"}. This record does not assert independent reproduction. ### Vanilla ASR: 41. Result ID: 69cf66a047f34ce924b8bc6c. [Reporting paper]() Recorded evaluation configuration: {"Target LLM":"Llama-3.1-8B"}. This record does not assert independent reproduction. ### TAP ASR: 37. Result ID: 69cf66a047f34ce924b8bc7a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"Sonnet-4"}. This record does not assert independent reproduction. ### AutoDAN-Turbo ASR: 33. Result ID: 69cf66a047f34ce924b8bc88. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"Haiku-3.5"}. This record does not assert independent reproduction. ### AutoDAN-Turbo ASR: 31. Result ID: 69cf66a047f34ce924b8bc89. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"Sonnet-4"}. This record does not assert independent reproduction. ### PAIR ASR: 29. Result ID: 69cf66a047f34ce924b8bc75. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"Sonnet-4"}. This record does not assert independent reproduction. ### PAP ASR: 20. Result ID: 69cf66a047f34ce924b8bc7f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"Sonnet-4"}. This record does not assert independent reproduction. ### SeqAR ASR: 16. Result ID: 69cf66a047f34ce924b8bc84. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"Sonnet-4"}. This record does not assert independent reproduction. ### SeqAR ASR: 9. Result ID: 69cf66a047f34ce924b8bc83. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"Haiku-3.5"}. This record does not assert independent reproduction. ### Vanilla ASR: 3. Result ID: 69cf66a047f34ce924b8bc6d. [Reporting paper]() Recorded evaluation configuration: {"Target LLM":"GPT-4o-mini"}. This record does not assert independent reproduction. ### Vanilla ASR: 3. Result ID: 69cf66a047f34ce924b8bc70. [Reporting paper]() Recorded evaluation configuration: {"Target LLM":"Sonnet-4"}. This record does not assert independent reproduction. ### Vanilla ASR: 2. Result ID: 69cf66a047f34ce924b8bc6e. [Reporting paper]() Recorded evaluation configuration: {"Target LLM":"GPT-4o"}. This record does not assert independent reproduction. ### Vanilla ASR: 1. Result ID: 69cf66a047f34ce924b8bc6f. [Reporting paper]() Recorded evaluation configuration: {"Target LLM":"Haiku-3.5"}. This record does not assert independent reproduction. ### SeqAR ASR: 0. Result ID: 69cf66a047f34ce924b8bc81. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"GPT-4o-mini"}. This record does not assert independent reproduction. ### SeqAR ASR: 0. Result ID: 69cf66a047f34ce924b8bc82. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Target LLM":"GPT-4o"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.