用本地化智能体自动发现新材料，成本低且结果媲美商业系统。
Hierarchical Deep Research with Local-Web RAG: Toward Automated System-Level Materials Discovery
- 构建分层研究树动态扩展与修剪科研路径，提升探索深度与连贯性。
- 在27个纳米材料课题中表现接近甚至超越商用系统，干实验验证可行。
- 支持本地部署，可接入私有数据与计算工具，适合科研机构使用。
我们提出一种面向复杂材料与器件发现的长周期、分层深度研究（DR）智能体，突破现有机器学习代理和闭源商业系统的局限。该框架部署于本地，结合本地检索增强生成与大语言模型推理，并引入深度研究树（DToR）机制，自适应扩展与剪枝研究分支，以最大化覆盖范围、深度与连贯性。我们在27个纳米材料/器件主题上进行系统评估，采用大语言模型作为评分裁判，五种主流网络可用模型作为评审员。此外，对五个代表性任务开展干实验验证，由领域专家利用密度泛函理论（DFT）等模拟工具检验智能体提案的可操作性。结果表明，该DR智能体生成报告的质量与商业系统（ChatGPT-5-thinking/o3/o4-mini-high Deep Research）相当甚至更优，且成本显著更低，同时支持本地数据与工具集成。
原文摘要 · Abstract (English)
We present a long-horizon, hierarchical deep research (DR) agent designed for complex materials and device discovery problems that exceed the scope of existing Machine Learning (ML) surrogates and closed-source commercial agents. Our framework instantiates a locally deployable DR instance that integrates local retrieval-augmented generation with large language model reasoners, enhanced by a Deep Tree of Research (DToR) mechanism that adaptively expands and prunes research branches to maximize coverage, depth, and coherence. We systematically evaluate across 27 nanomaterials/device topics using a large language model (LLM)-as-judge rubric with five web-enabled state-of-the-art models as jurors. In addition, we conduct dry-lab validations on five representative tasks, where human experts use domain simulations (e.g., density functional theory, DFT) to verify whether DR-agent proposals are actionable. Results show that our DR agent produces reports with quality comparable to--and often exceeding--those of commercial systems (ChatGPT-5-thinking/o3/o4-mini-high Deep Research) at a substantially lower cost, while enabling on-prem integration with local data and tools.
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。