协议引导的 AI 网站冗余审计：从主观判断到可审计标准
原标题：From Subjective Judgments to Auditable Standards:Protocol-Guided AI Auditing of Website Redundancy
AI 摘要
arXiv 论文提出 Cora（反事实可观察冗余审计）框架，用于可审计的网站冗余 AI 审计。Cora 将冗余分解为重复负载、正常使用税和故障域恢复储备三个维度，并通过版本化视觉语言模型提出注释，经类型化验证和发布检查后选择性发布分数。在透明机制测试台上，Cora 的分解表示优于标量负载基线，但两个小型视觉语言模型因未满足发布要求而被拒绝自动评分。该框架定位为受控基准下的候选审计程序，而非通用标准。
正文节选
From Subjective Judgments to Auditable Standards: Protocol-Guided AI Auditing of Website Redundancy Abstract Website redundancy does not have a single fixed meaning. The same repeated element may distract during one task and provide backup during another. We introduce Cora (Counterfactual, Observable Redundancy Audit), which measures repetition load, normal-use tax, and failure-domain recovery reserve separately. Each run retains screenshots, stable element identities, and task traces. A version