# Structural Erasure on WMDP bio Source: [Structural Erasure on WMDP bio]() Collection index: [llms.txt]() Data updated: 2026-05-15T09:59:37.182Z Evaluation of Structural Erasure performance on the WMDP bio dataset, reporting metrics including CAD, AS-C, and AS-NC. 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 | Structural Erasure | | Dataset (including any stated split) | WMDP bio | | Metric | CAD | | Unit | SCALAR | | Direction | Lower 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 - [CAD](): Lower is better; unit: SCALAR. - [AS-C](): Lower is better; unit: SCALAR. - [AS-NC](): Lower is better; unit: SCALAR. ## Recorded results | Method | CAD | Result date | Report paper | | --- | --- | --- | --- | | Zero-shot | 0 | Not recorded | 6a06a8c9dbc836b87d82a365 | | Zero-shot | 0 | Not recorded | 6a06a8c9dbc836b87d82a365 | | LUNAR | 0.039 | Not recorded | 6a06a8c9dbc836b87d82a365 | | LUNAR | 0.041 | Not recorded | 6a06a8c9dbc836b87d82a365 | | Surgical GA | 0.321 | Not recorded | 6a06a8c9dbc836b87d82a365 | | Global GA | 0.523 | Not recorded | 6a06a8c9dbc836b87d82a365 | | Surgical GA | 0.663 | Not recorded | 6a06a8c9dbc836b87d82a365 | | GU+SimNPO | 0.792 | Not recorded | 6a06a8c9dbc836b87d82a365 | | NPO | 0.847 | Not recorded | 6a06a8c9dbc836b87d82a365 | | NPO | 0.849 | Not recorded | 6a06a8c9dbc836b87d82a365 | | SimNPO | 0.911 | Not recorded | 6a06a8c9dbc836b87d82a365 | | Global GA | 1.021 | Not recorded | 6a06a8c9dbc836b87d82a365 | | MANSU | 1.089 | Not recorded | 6a06a8c9dbc836b87d82a365 | | MANSU | 1.143 | Not recorded | 6a06a8c9dbc836b87d82a365 | | GU+SimNPO | 1.292 | Not recorded | 6a06a8c9dbc836b87d82a365 | | SimNPO | 1.433 | Not recorded | 6a06a8c9dbc836b87d82a365 | ### Zero-shot CAD: 0. Result ID: 6a06ee7247f34ce924ef1bcf. [Reporting paper]() Recorded evaluation configuration: {"Model":"Llama-3.1-8B-Instruct"}. This record does not assert independent reproduction. ### Zero-shot CAD: 0. Result ID: 6a06ee7247f34ce924ef1bd7. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen-3-8B"}. This record does not assert independent reproduction. ### LUNAR CAD: 0.039. Result ID: 6a06ee7247f34ce924ef1bdd. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen-3-8B"}. This record does not assert independent reproduction. ### LUNAR CAD: 0.041. Result ID: 6a06ee7247f34ce924ef1bd5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama-3.1-8B-Instruct"}. This record does not assert independent reproduction. ### Surgical GA CAD: 0.321. Result ID: 6a06ee7247f34ce924ef1bd1. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Llama-3.1-8B-Instruct"}. This record does not assert independent reproduction. ### Global GA CAD: 0.523. Result ID: 6a06ee7247f34ce924ef1bd0. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Llama-3.1-8B-Instruct"}. This record does not assert independent reproduction. ### Surgical GA CAD: 0.663. Result ID: 6a06ee7247f34ce924ef1bd9. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Qwen-3-8B"}. This record does not assert independent reproduction. ### GU+SimNPO CAD: 0.792. Result ID: 6a06ee7247f34ce924ef1bdc. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen-3-8B"}. This record does not assert independent reproduction. ### NPO CAD: 0.847. Result ID: 6a06ee7247f34ce924ef1bda. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen-3-8B"}. This record does not assert independent reproduction. ### NPO CAD: 0.849. Result ID: 6a06ee7247f34ce924ef1bd2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama-3.1-8B-Instruct"}. This record does not assert independent reproduction. ### SimNPO CAD: 0.911. Result ID: 6a06ee7247f34ce924ef1bdb. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen-3-8B"}. This record does not assert independent reproduction. ### Global GA CAD: 1.021. Result ID: 6a06ee7247f34ce924ef1bd8. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Qwen-3-8B"}. This record does not assert independent reproduction. ### MANSU CAD: 1.089. Result ID: 6a06ee7247f34ce924ef1bde. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen-3-8B"}. This record does not assert independent reproduction. ### MANSU CAD: 1.143. Result ID: 6a06ee7247f34ce924ef1bd6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama-3.1-8B-Instruct"}. This record does not assert independent reproduction. ### GU+SimNPO CAD: 1.292. Result ID: 6a06ee7247f34ce924ef1bd4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama-3.1-8B-Instruct"}. This record does not assert independent reproduction. ### SimNPO CAD: 1.433. Result ID: 6a06ee7247f34ce924ef1bd3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama-3.1-8B-Instruct"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.