开源智能体平台AGAPI提升材料设计精度，工具调用显著降低预测误差。
AGAPI-Agents: An Open-Access Agentic AI Platform for Accelerated Materials Design on AtomGPT.org
- 整合8个LLM与28个科学工具，实现自主材料分析
- 工具增强后平均绝对误差低于0.005 eV，无工具时达1.25–1.86 eV
- 支持多步缺陷工程自动化，适合材料科研与算法开发者
Agentic AI系统将大语言模型与外部科学工具连接，但工具使用是否提升预测精度尚不明确。我们推出AGAPI（AtomGPT.org API），一个开源平台，集成8个开源大语言模型与18个REST接口（28个代理工具、50个网页应用），覆盖材料数据库、力场、紧束缚能带、X射线衍射及蛋白质结构。基于JARVIS-Leaderboard电子结构测试集的三重评估残差分解，分离出代理流程保真度与继承的密度泛函理论（DFT）功能偏差。对于体模量和带隙，代理可精确复现JARVIS-DFT结果，实验参考降级源于功能偏差而非代理故障。在抗记忆测试集（57个缺陷超胞、60种假设组分）中，工具增强后的平均绝对误差（MAE）低于0.005 eV，而无工具时为1.25至1.86 eV，证实工具在参数知识缺失场景下不可或缺。进一步展示包含10步操作的自主多步工作流，如缺陷工程。AGAPI可在https://github.com/atomgptlab/agapi获取。
原文摘要 · Abstract (English)
Agentic AI systems increasingly connect large language models (LLMs) to external scientific tools, yet whether and when tool access improves prediction accuracy remains uncharacterized. We present AGAPI (AtomGPT.org API), an open access platform integrating eight open-source LLMs with 18 REST endpoints (28 agent tools, 50 web apps) spanning materials databases, force fields, tight-binding band structures, X-ray diffraction, and protein structure. A three-evaluation residual decomposition on JARVIS-Leaderboard electronic-structure test sets separates agent pipeline fidelity from inherited density functional theory (DFT) functional bias. For bulk modulus and bandgap the agent reproduces JARVIS-DFT entries to numerical precision, so the experimental-reference degradation is functional bias, not agentic malfunction. On memorization-resistant test sets (57 defective supercells, 60 hypothetical compositions), tool-augmented mean absolute error (MAE) is below 0.005 eV versus 1.25 to 1.86 eV tool-free, confirming tools are indispensable where parametric knowledge is unavailable. We further demonstrate autonomous multi-step workflows including 10-operation defect-engineering pipelines. AGAPI is available at https://github.com/atomgptlab/agapi.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。