用确定性引擎验证语言模型设计，确保物理可靠性。
Structural Certification for Reliable Physical Design with Language Models
- 模型提方案，独立引擎判定真伪，防伪造
- 80次对抗测试零误认证，即使引擎有缺陷
- 适合需要高可靠性的工程设计场景
若将断言权限从语言模型转移至确定性引擎，即可使不可靠的模型生成可靠的物理设计：模型提出方案，引擎独立认证，返回已认证、不可能或未知。我们提出物理锚定认证（PHACT）框架，覆盖五个科学领域，明确可信认证的关键条件。仅接受模型输出的检查器可被伪造；而通过固定输入直接推导认证结果，则从构造上杜绝伪造可能。在涵盖两个模型、两种解码温度及一个故意故障引擎的80次对抗测试中，该机制实现零误认证。
原文摘要 · Abstract (English)
An unreliable language model can be made to produce reliable physical designs if the authority to assert is moved out of the model: the model proposes, and a deterministic engine alone certifies, returning certified, impossible, or unknown. We introduce Physics-Anchored Certification (PHACT), a propose-certify loop spanning five scientific domains, and identify what makes such a certificate trustworthy. A checker that accepts a model-supplied value can be forged; deriving the certified quantity from fixed inputs instead makes forgery impossible by construction. Across eighty adversarial trials spanning two models, two decoding temperatures, and a deliberately faulted engine, this contract produced zero false certifications.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。