让AI复现实验，成功率从76%到93%
AI 复现别人的机器学习实验，常常卡在「论文没写全」的细节上。这篇的发现是：把这些散落的操作知识整理成一份标准「复现包」，连能力较弱的模型也能把成功率从 76.4% 拉到 92.6%，还省下近两成 API 开销。他们做的不是更强的模型，而是一套固定的产物结构——精修过的代码库、环境配置、实验启动脚本，外加一份复现指南，再配一个专门核对执行证据的「审计员」，让智能体在干净会话里照着跑。四个模型之间的成功率差距从 21.6 个百分点缩到 3.9，全部越过 90%。拆掉任何一块都会掉分，代码和环境缺一不可。它不是你明天能直接用的工具，但它指出一个方向：让 AI 干活，喂给它「怎么做的记录」比换更大的模型更划算。
📄 原文摘要(英文)
Reproducing machine learning experiments requires reconstructing procedures from information scattered across papers, released code, and supporting artifacts. For coding agents, reproduction success can therefore depend on their ability to fill in missing details, rather than on the released artifacts alone. We ask whether preserving recovered procedures and decisions improves subsequent independent reproduction. We define a standardized artifact design comprising a refined codebase, environment setup, experiment launchers, and a reproduction guide, and introduce ReproBridge, an agent workflow for constructing these refined reproducible artifacts. ReproBridge-Bench provides 1,029 tasks from 387 experiments across 36 papers, anchored by five human-verified reference scores per task. Agents attempt tasks in fresh sessions with raw code or refined artifacts; a separate Auditor checks execution evidence. Across four evaluated models, refined artifacts raise mean reproduction success from 76.4% to 92.6% and reduce downstream API spending by 19.7% on average. The cross-model success-rate spread narrows from 21.6 to 3.9 percentage points, with all four models achieving mean success rates above 90%. Removing any component reduces success in the ablation study, with the largest losses from code and environment removal. These results suggest that preserving execution knowledge helps even models with lower raw-code success reproduce experiments more reliably and at lower API cost.