# Jailbreaking on AdvBench (WASR & W-Toxicity Score) Source: [Jailbreaking on AdvBench (WASR & W-Toxicity Score)]() Collection index: [llms.txt]() Data updated: 2026-05-19T09:31:56.596Z Evaluation of model susceptibility to jailbreaking prompts using the AdvBench dataset, reporting Weighted Attack Success Rate (WASR) and Weighted Toxicity Score. 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) | AdvBench | | Metric | WASR | | 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 - [WASR](): Higher is better; unit: PERCENT. - [W-Toxicity Score](): Lower is better; unit: PERCENT. ## Recorded results | Method | WASR | Result date | Report paper | | --- | --- | --- | --- | | wide-net-casting jailbreak method (Ours) | 73.3 | Not recorded | 6a0beec1dbc836b87d82c63a | | Naive Strategy 2 | 56.2 | Not recorded | 6a0beec1dbc836b87d82c63a | | Naive Strategy 1 | 55.5 | Not recorded | 6a0beec1dbc836b87d82c63a | | wide-net-casting jailbreak method (Ours) | 51.9 | Not recorded | 6a0beec1dbc836b87d82c63a | | ReMiss | 50.2 | Not recorded | 6a0beec1dbc836b87d82c63a | | wide-net-casting jailbreak method (Ours) | 40.6 | Not recorded | 6a0beec1dbc836b87d82c63a | | Naive Strategy 2 | 38 | Not recorded | 6a0beec1dbc836b87d82c63a | | Naive Strategy 1 | 37.3 | Not recorded | 6a0beec1dbc836b87d82c63a | | ReMiss | 33.1 | Not recorded | 6a0beec1dbc836b87d82c63a | | Naive Strategy 2 | 31.8 | Not recorded | 6a0beec1dbc836b87d82c63a | | Naive Strategy 1 | 31.1 | Not recorded | 6a0beec1dbc836b87d82c63a | | ReMiss | 28.5 | Not recorded | 6a0beec1dbc836b87d82c63a | ### wide-net-casting jailbreak method (Ours) WASR: 73.3. Result ID: 6a0c2d63fbc373592c029570. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Safety Alignment":"Original Safety Alignment"}. This record does not assert independent reproduction. ### Naive Strategy 2 WASR: 56.2. Result ID: 6a0c2d63fbc373592c02956d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Safety Alignment":"Original Safety Alignment"}. This record does not assert independent reproduction. ### Naive Strategy 1 WASR: 55.5. Result ID: 6a0c2d63fbc373592c02956a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Safety Alignment":"Original Safety Alignment"}. This record does not assert independent reproduction. ### wide-net-casting jailbreak method (Ours) WASR: 51.9. Result ID: 6a0c2d63fbc373592c029571. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Safety Alignment":"Original Safety Alignment + SmoothLLM"}. This record does not assert independent reproduction. ### ReMiss WASR: 50.2. Result ID: 6a0c2d63fbc373592c029567. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Safety Alignment":"Original Safety Alignment"}. This record does not assert independent reproduction. ### wide-net-casting jailbreak method (Ours) WASR: 40.6. Result ID: 6a0c2d63fbc373592c029572. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Safety Alignment":"Original Safety Alignment + RobustKV"}. This record does not assert independent reproduction. ### Naive Strategy 2 WASR: 38. Result ID: 6a0c2d63fbc373592c02956e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Safety Alignment":"Original Safety Alignment + SmoothLLM"}. This record does not assert independent reproduction. ### Naive Strategy 1 WASR: 37.3. Result ID: 6a0c2d63fbc373592c02956b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Safety Alignment":"Original Safety Alignment + SmoothLLM"}. This record does not assert independent reproduction. ### ReMiss WASR: 33.1. Result ID: 6a0c2d63fbc373592c029568. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Safety Alignment":"Original Safety Alignment + SmoothLLM"}. This record does not assert independent reproduction. ### Naive Strategy 2 WASR: 31.8. Result ID: 6a0c2d63fbc373592c02956f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Safety Alignment":"Original Safety Alignment + RobustKV"}. This record does not assert independent reproduction. ### Naive Strategy 1 WASR: 31.1. Result ID: 6a0c2d63fbc373592c02956c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Safety Alignment":"Original Safety Alignment + RobustKV"}. This record does not assert independent reproduction. ### ReMiss WASR: 28.5. Result ID: 6a0c2d63fbc373592c029569. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Safety Alignment":"Original Safety Alignment + RobustKV"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.