利用智能体记忆构建面向材料科学家的终身AI合作伙伴
Harnessing agent memory to build lifelong AI partners for materials scientists
浏览论文内容
中文总结 AI 辅助
该研究提出一种自进化记忆框架，可跨模型迁移材料研究经验，使GPT-5.2任务成功率翻倍，减少重复错误，降低材料模拟工作流程的追踪负担与工具调用次数。
中文摘要 AI 辅助
材料研究的进展依赖于积累的经验——有效的脚本、可信的方案、失败计算或实验附带的警告，以及将新问题与旧结果关联起来的判断。这些经验对于可重复性和知识转移至关重要，但通常分散在笔记本、存储库、作业日志和个人记忆中，且很少能在人工智能智能体之间迁移。本文提出，可围绕持久记忆而非特定智能体实现方式，设计面向材料科学的终身AI合作伙伴。我们引入一种自进化记忆框架，将科学经验存储为可检查的事实和可执行的技能，从而能跨模型检索、修改和迁移观测结果、失败边界、方案及验证检查。我们在三个计算场景中评估该理念，这些场景展现了材料研究能力的不同层面：在包含138个可执行子任务的49个真实材料工具使用问题中，记忆使GPT-5.2的任务成功率几乎翻倍，且无需更新模型参数；在元素固体的物态方程计算中，记忆将波函数初始化失败转化为执行前的防护措施，使正确/部分/错误结果从22/1/4提升至25/2/0，并避免了92%的重复错误；在13个实用材料模拟工作流程中，记忆的技能和失败事实使总追踪负担（token）减半，到第三轮时工具调用减少了一倍以上，同时在带隙、声子、空位和功函数分析中保留了具有物理意义的输出。这些结果表明，智能体记忆可作为一种持久的科学资产，是一份可移植、自我改进的材料研究经验记录，其生命周期长于任何单一模型或智能体栈。
英文摘要
Materials research advances through accumulated experience - scripts that work, protocols that are trusted, warnings attached to failed calculations or experiments, and judgement that links a new question to an old result. This experience is essential for reproducibility and knowledge transfer, yet it is usually fragmented across notebooks, repositories, job logs and individual memory, and it is rarely portable across artificial-intelligence agents. Here we argue that a lifelong AI partner for materials science can be designed around persistent memory rather than around a particular agent implementation. We introduce a self-evolving memory framework that stores scientific experience as inspectable facts and executable skills, so that observations, failure boundaries, protocols and validation checks can be retrieved, revised and migrated across models. We evaluate the idea in three computational settings that expose different layers of materials-research competence. In 49 real-world materials-tool-use questions comprising 138 executable subtasks, memory nearly doubles GPT-5.2 task success without model-parameter updates. In elemental-solid equation-of-state calculations, memory converts a wavefunction-initialization failure into a pre-execution guardrail, improving outcomes from 22/1/4 to 25/2/0 Correct/Partial/Error and avoiding 92% of repeated errors. In 13 practical material simulation workflows, remembered skills and failure facts halve the aggregate trace burden (tokens) and reduce tool calls by over a factor of two by the third round, while preserving physically meaningful outputs in band-gap, phonon, vacancy and work-function analyses. These results show that agent memory can serve as a durable scientific asset; a portable, self-improving record of materials-research experience that outlives any single model or agent stack.
发表机构
- The University of Hong Kong（香港大学）
- Materials Innovation Institute for Life Sciences and Energy (MILES), HKU-SIRI（香港大学-深圳国际研究院生命科学与能源材料创新研究院）
- International Digital Economy Academy (IDEA)（国际数字经济学院）
机构由 AI 辅助整理，请以论文原文为准。
