发表机构
Neapolis University, Pafos, Cyprus; Constructor Labs, Bremen, Campus Ring˜1, 28759, Germany; Constructor University, Bremen, Campus Ring˜1, 28759, Germany; Institute for Functional Intelligent Materials, National University of Singapore, 4 Science Drive˜2, Singapore 117544, Singapore（尼波利斯大学; 康斯特实验室; 康斯特大学; 新加坡国立大学功能智能材料研究所）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究连接GitHub和arXiv上的AI研究智能体注册表，检验设计规律，揭示论文与代码库关联的策展性质，并明确公共可见性主张的局限。
AI 中文摘要
AI研究智能体，或称自动研究系统，将语言模型与工具、搜索、评估以及研究产物的迭代修改相结合。其公开软件生态难以比较，因为代码库、论文、基准、库和配套产物常常被计为同一群体。我们连接了两个由智能体起草、人工裁决的注册表，冻结于2026年6月10日：139个规范公开代码库和101篇关于AI辅助研究系统的论文，每条记录都附有锚定于其原始来源的证据卡。九个被推广的设计谱系包含52个代码库中的59个成员资格，其中七个被分配到两个谱系。基于这些证据，我们检验了六个带时间戳的候选设计规律。对25个新纳入代码库的前瞻性审计观察到六个指定触发事件中的零个，为R1-R4提供了有限支持；更广泛的语料证据缩小了R5的原始范围，而一个失败的方向性预测使R6保持探索性。论文-代码库图包含212条由策展人分配的相关性边：101篇论文中有64篇至少携带一条边，这些边触及139个代码库中的23个，而六个关联最多的代码库承载了212条边中的120条（56.6%）。对七篇论文的全文检查表明，该关系衡量的是我们的策展，而非论文的引用行为。一项针对提交历史和公开贡献者记录裁决的身份审计确认，至少有18位论文作者同时也是规范注册表代码库的作者（在13篇论文的63个候选作者位置中占20个）；一个按署名匹配的层级达到58个身份，但单独报告且证据较弱。本研究贡献了一个在预期测试下的源基础理论、一个由策展人分配的论文-代码库相关性图谱，以及对关于公开可见性和社区结构的主张的明确限制。
英文摘要
AI-research agents, or autoresearch systems, combine language models with tools, search, evaluation, and iterative modification of research artifacts. Their public software ecology is hard to compare because repositories, papers, benchmarks, libraries, and companion artifacts are often counted as one population. We connect two agent-drafted, human-adjudicated registries frozen on 10 June 2026: 139 canonical public repositories and 101 papers on AI-assisted research systems, each record carrying an evidence card anchored to its primary sources. Nine promoted design lineages contain 59 memberships among 52 repositories, seven of them assigned to two lineages. Against this evidence we examine six time-stamped candidate design regularities. A prospective audit of 25 newly ingested repositories observed zero of six specified trigger events, supplying bounded support for R1-R4; broader corpus evidence narrowed the original scope of R5, and a failed directional prediction leaves R6 exploratory. The paper-repository graph contains 212 curator-assigned relatedness edges: 64 of 101 papers carry at least one edge, the edges reach 23 of 139 repositories, and the six most-linked repositories carry 120 of the 212 edges (56.6%). A full-text check of seven papers shows the relation measures our curation, not the papers' citation behavior. An identity audit adjudicated against commit histories and public contributor records confirms at least 18 paper authors who also author a canonical registry repository (20 of 63 candidate author slots in 13 of 101 papers); a byline-matched tier reaches 58 identities but is reported separately and more weakly. The study contributes a source-grounded theory under prospective test, a curator-assigned map of paper-repository relatedness, and explicit limits on claims about public visibility and community structure.
Comments56 pages, 1 figure
