发表机构
UCLA; MIT; Amazon; UPenn（加利福尼亚大学洛杉矶分校; 麻省理工学院; 亚马逊公司; 宾夕法尼亚大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
研究多步逆合成规划难题，提出RetroAgent智能体，通过结合结构化记忆桥接符号搜索与神经推理，利用记忆和化学工具观察搜索状态，实验证明其在分布内和分布外基准测试中性能强且泛化能力好。
AI 中文摘要
多步逆合成规划旨在通过一系列可行反应将目标分子分解为市售的构建模块。巨大的组合搜索空间使这项任务即使对专业化学家来说也具有挑战性。传统方法将树搜索与离线训练的价值网络相结合，孤立地对候选物进行评分，而不考虑完整的多步路线。最近的工作利用大语言模型来完成这项任务，但依赖于简单的接口，限制了对整个搜索空间的探索。我们引入了RetroAgent，这是一种大语言模型智能体，它通过与结构化记忆的结合来桥接符号搜索和神经推理。通过记忆和化学工具，智能体观察完整的搜索状态，包括探索过的路线、可用的替代方案和中间体的性质，从而基于全局进展和领域知识做出明智的决策。在分布内和分布外基准上的实验表明，RetroAgent具有强大的性能和泛化能力。
英文摘要
Multi-step retrosynthesis planning seeks to decompose a target molecule into commercially available building blocks through a sequence of feasible reactions. The vast combinatorial search space makes this task challenging even for expert chemists. Traditional methods combine tree search with offline-trained value networks that score candidates in isolation, without reasoning about complete multi-step routes. Recent work leverages Large Language Models (LLMs) for this task, but relies on simple interfaces that limit exploration of the full search space. We introduce RetroAgent, an LLM agent that bridges symbolic search and neural reasoning through a harness with structured memory. Through memory and chemistry tools, the agent observes the full search state, including explored routes, available alternatives, and properties of intermediates, enabling informed decisions grounded in both global progress and domain knowledge. Experiments on in-distribution and out-of-distribution benchmarks demonstrate that RetroAgent delivers strong performance and generalization.
CommentsTo appear at COLM 2026
