arXiv cs.MA· Juan Diego Toscano, Zhaojie Chai, George Em Karniadakis·· 12 小时前AI 评分38
GRAFT-ATHENA：面向自主发现与进化数值算法的自我改进智能体团队
GRAFT-ATHENA: Self-Improving Agentic Teams for Autonomous Discovery and Evolutionary Numerical Algorithms
AI 导读
研究者提出 GRAFT（Graph Reduction to Adaptive Factored Trees），将语言模型智能体在科学工作流中的问题—方法映射显式化为可扩展的概率结构，并据此构建自我改进的智能体团队 GRAFT-ATHENA。
正文 · 原文
Abstract:Language-model agents plan, execute, and evaluate scientific workflows. However, the problem--method relationships guiding their choices remain implicit, so plausible outputs can violate the assumptions of the scientific problem. We introduce Graph Reduction to Adaptive Factored Trees (GRAFT), which makes this problem-to-method map explicit as an expandable probabilistic structure of admissible problems, methods, and dependencies. Because this structure follows scientific semantics, knowledge from documentation, papers, and domain experts can be encoded, while factorization keeps the structure tractable. Semantic fingerprints of encoded problems measure their similarity, so expertise and experience guide related problems. GRAFT-guided workflows exceeded published expert baselines in physics-informed learning, recovered blood-rheology trends, predicted Apollo Command Module hypersonic loading within 1.8\% of experiment, and established machine-checked error bounds for regularized brain-flow reconstruction and a spectral physics-informed architecture. The representation retains evaluated outcomes and incorporates new methods, allowing scientific knowledge and experience to accumulate.
| Subjects: | Machine Learning (cs.LG); Multiagent Systems (cs.MA); Probability (math.PR) |
| Cite as: | arXiv:2605.11117 [cs.LG] |
| (or arXiv:2605.11117v3 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2605.11117 arXiv-issued DOI via DataCite |
Submission history
From: Juan Toscano [view email]
[v1]
Mon, 11 May 2026 18:27:01 UTC (15,141 KB)
[v2]
Tue, 15 Sep 2026 01:32:10 UTC (12,182 KB)
[v3]
Fri, 16 Oct 2026 17:10:34 UTC (16,753 KB)
来源：arXiv cs.MA · arxiv.org