AI for Science Multi-Agent Collaboration Autonomous Planning Scientific Discovery
摘要

科学发现需要在广阔搜索空间中具备智能、毅力与偶然性。当前顶尖科学能力彼此孤立,无法预设团队以应对所有需求。Science Earth 是一个行星级科学运行时,通过 EACN 协议让仿真集群、湿实验室机器人等能力自主发现、协商任务归属并裁决证据标准,无需预先设计工作流。实验表明,该机制能在跨太平洋同步性研究及单细胞分析中,自发形成协作结构,纠正理论假设并产出新成果,推动科学推理成为分布式自修正过程。

AI 推荐理由

论文核心在于构建动态协作架构,使 Agent 能自主协商任务归属并生成 emergent 协作结构,属高级规划。

研究机构
Department of Pathology, Department of Genetics, Stanford University School of Medicine, Stanford, CA, USA Princeton AI Lab, Department of Electrical & Computer Engineering, Princeton University, Princeton, NJ, USA Scripps Research, La Jolla, CA, USA Division of Biostatistics, Department of Population Health, New York University Grossman School of Medicine, New York, NY, USA College of Computing and Data Science, Nanyang Technological University, Singapore Department of Computer Science, Yale University, New Haven, CT, USA Department of Physics, Princeton University, Princeton, NJ, USA
论文信息
作者 Zhe Zhao, Haibin Wen, Yingcheng Wu, Jiaming Ma, Yifan Wen et al.
发布日期 2026-05-31
arXiv ID 2606.01316
相关性评分 9/10 (高度相关)