AIM：用于自动化研究的智能体式想法管理

中文摘要 AI 辅助

前沿大语言模型（LLMs）越来越多地通过迭代搜索被用于自动化科学研究。我们区分了想法驱动型搜索与解决方案驱动型搜索，并识别出三个核心挑战：组织不断演进的研究想法、选择有前景的方向，以及维持想法与其实现之间的一致性。为应对这些挑战，我们引入了智能体式想法管理器（Agentic Idea Manager，AIM），这是一个完全自主的框架，用于在想法驱动的自动化研究中管理和探索研究方向。受贝叶斯优化的启发，AIM使用智能体式代理（Agentic Surrogate）和智能体式采集（Agentic Acquisition）机制来组织已发现的想法并指导其选择。一个解决方案审计器（Solution Auditor）维护想法-解决方案的完整性，而资源规划器（Resource Planner）在并行搜索分支之间自适应地分配剩余的实验预算。在10个AutoLab基准任务上的实验表明，AIM在系统优化（System Optimization）任务上超过最强基线1.6个百分点，在长时程模型开发与CUDA（Model Development & CUDA）任务上超过4.9个百分点。值得注意的是，AIM达到最佳基线性能的速度在墙钟时间上最多快3.1倍。我们进一步提供了关于何时在想法上进行搜索有益的理论分析。我们的分析表明，显式的想法级分配使得语义覆盖直接可控，并且当竞争性研究方向在众多可行备选中较为稀疏时，更广泛的覆盖变得更有价值。项目页面：此https URL

英文摘要

Frontier LLMs are increasingly used to automate scientific research through iterative search. We distinguish idea-driven search from solution-driven search and identify three core challenges: organizing evolving research ideas, selecting promising directions, and maintaining alignment between ideas and their implementations. To address these challenges, we introduce the Agentic Idea Manager (AIM), a fully autonomous framework for managing and exploring research directions in idea-driven automated research. Inspired by Bayesian optimization, AIM uses an Agentic Surrogate and an Agentic Acquisition mechanism to organize discovered ideas and guide their selection. A Solution Auditor maintains idea-solution integrity, while a Resource Planner adaptively allocates the remaining experimental budget across parallel search branches. Experiments on 10 AutoLab benchmark tasks show that AIM surpasses the strongest baseline by 1.6 percentage points on System Optimization tasks and 4.9 percentage points on long-horizon Model Development & CUDA tasks. Notably, AIM reaches the best baseline performance up to 3.1x faster in wall-clock time. We further provide a theoretical analysis of when searching over ideas becomes beneficial. Our analysis shows that explicit idea-level allocation makes semantic coverage directly controllable, and that broader coverage becomes increasingly valuable when competitive research directions are sparse among many plausible alternatives. Project Page: https://imhgchoi.github.io/agentic-idea-manager/

发表机构

• Google Cloud AI Research（谷歌云AI研究）
• University of Wisconsin-Madison（威斯康星大学麦迪逊分校）

机构由 AI 辅助整理，请以论文原文为准。