证据门控研究：自适应搜索中的统计受控模型采纳
Evidence-Gated Research: Statistically Controlled Model Adoption in Adaptive Search
浏览论文内容
中文总结 AI 辅助
提出证据门控研究（EGR）作为自适应搜索中模型采纳的统计控制层，通过冻结挑战者并构建随时有效证据来控制错误发现率，实验表明其比现有方法节省56.1%的决策证据且无持久虚假采纳。
中文摘要 AI 辅助
自适应模型搜索具有路径依赖性：一旦某个挑战者被采纳，它就成为后续候选模型生成的参照。因此，一个统计上不支持替换的模型可能会改变尚未提出的假设。我们提出了证据门控研究（EGR），一种用于移动在位者搜索的统计采纳层。EGR在决策证据揭示之前冻结每个挑战者，在预先声明的一组环境中构建随时有效的证据，将证据可预测地引导至未解决的组件，组合一个持久的候选e值，并将该e值传递给在线控制器。在明确的条件有效性和可预测性条件下，即使早期采纳改变了后续挑战者，所得到的程序也能控制已声明的全环境采纳目标的错误发现率。在一个包含5,000条轨迹的闭环基准测试中，仅开发阶段的e-LOND达到了持久FDR 0.621，而在该有限运行中，经过审计的EGR变体未观察到持久的虚假采纳路径。在600对挑战者-在位者对的匹配重放中，阶段式EGR保留了固定的随时替代交叉决策，同时在代表性阈值下使用了56.1%更少的决策证据。一项三环境公共数据研究和一项包含40,000个样本的受控神经基准测试重现了证据效率模式。这些结果将模型替换确定为自适应模型开发中一个独特的统计控制点。
英文摘要
Adaptive model search is path dependent: once a challenger is adopted, it becomes the reference from which later candidates are generated. A statistically unsupported replacement can therefore alter hypotheses that have not yet been proposed. We introduce Evidence-Gated Research (EGR), a statistical adoption layer for moving-incumbent search. EGR freezes each challenger before decision evidence is revealed, builds anytime-valid evidence across a predeclared set of environments, routes evidence predictably toward unresolved components, composes a persistent candidate e-value, and passes that e-value to an online controller. Under explicit conditional-validity and predictability conditions, the resulting procedure controls false discovery rate for the declared all-environment adoption target even though earlier adoptions change later challengers. In a 5,000-trajectory closed-loop benchmark, development-only e-LOND attains persistent FDR 0.621, whereas no persistent false-adoption path is observed for the audited EGR variants in that finite run. In matched replay over 600 challenger--incumbent pairs, Stagewise EGR preserves fixed-anytime alternative crossing decisions while using 56.1% less decision evidence at the representative threshold. A three-environment public-data study and a 40,000-sample controlled neural benchmark reproduce the evidence-efficiency pattern. These results identify model replacement as a distinct statistical control point in adaptive model development.
发表机构
- Stony Brook University（石溪大学）
机构由 AI 辅助整理，请以论文原文为准。
