跳到正文
arXiv cs.AI· Kaixin Li, Meng Luo, Yanlin Li, Yuchen Tian, Xiang Liu, Xiangru Jian, Yan Yang, Zhiheng Xi, Yang Zhang, Ziyang Luo, Mingfei Cheng, Hongzhan Lin·· 4 小时前AI 评分38

当世界成为测试套件:面向物理世界的智能体智能

When the World Is the Test Suite: Agentic Intelligence for the Physical World

AI 导读

一篇 83 页论文将物理世界视为程序的测试套件,系统梳理为机器人、实验室和工业控制器编写程序的编码智能体。论文围绕智能体如何为物理世界写程序、这些程序如何被测试、以及智能体如何从结果中学习三个问题展开,指出写程序正变得廉价,而可信的验证证据依然昂贵。作者提出进展将依赖物理版的持续集成:验证要便宜到能频繁运行、可靠到能在搜索中信任,并与它所验证的经验同步维护。

正文 · 原文

Authors:Kaixin Li, Meng Luo, Yanlin Li, Yuchen Tian, Xiang Liu, Xiangru Jian, Yan Yang, Zhiheng Xi, Yang Zhang, Ziyang Luo, Mingfei Cheng, Hongzhan Lin

View PDF HTML (experimental)

Abstract:Coding agents built to repair software now write, test, and repair programs for robots, laboratories, and industrial controllers. Unlike in software engineering, no test suite comes with the code. Whether a program works must be established by running it on a robot, in a physics simulator, or through the policies it helps to train, and such evidence is costly to obtain and can itself be wrong. As these agents take on more physical tasks, how their programs are judged matters as much as how they are written. This paper examines this emerging area, agentic intelligence for the physical world, by treating the world as the program's test suite. We organize the discussion around three questions: how agents write programs for the physical world, how these programs are tested, and how agents learn from the results. We first trace how language models for robots and software engineering agents came together, ask which software testing practices carry over to physical tests, and describe each program by where it sits relative to the actuator and by what judges it. For writing, we cover programs that act as robot policies and programs that support policy learning by producing rewards, tasks, and training worlds. For testing, we examine what physical trials cost, what traces, learned judges, simulators, and formal checks can establish in their place, and what safety requires beyond task success. For learning, we discuss how tested programs are retained as experience, in skill libraries and in model weights. Finally, we turn to benchmarks and evaluation practice and outline open challenges. Writing programs is becoming cheap, while trustworthy evidence about what they do remains costly. Progress will therefore depend on a physical counterpart of continuous integration: verification cheap enough to run often, reliable enough to trust under search, and maintained alongside the experience it validates.
Comments: 83 pages. Project page: this https URL. GitHub: this https URL
Subjects: Robotics (cs.RO); Artificial Intelligence (cs.AI); Software Engineering (cs.SE)
Cite as: arXiv:2611.22291 [cs.RO]
  (or arXiv:2611.22291v1 [cs.RO] for this version)
  https://doi.org/10.48550/arXiv.2611.22291

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Lin Hongzhan [view email]
[v1] Tue, 17 Nov 2026 06:48:03 UTC (12,050 KB)

来源:arXiv cs.AI · arxiv.org