cs.RO提交于 2026-10-30已译

Gym-Anything for Robotics：用数字智能体扩展物理仿真

Gym-Anything for Robotics: Scaling physical simulations with digital agents

Pranjal Aggarwal · Sean Welleck

中文摘要

逼真且快速的机器人仿真有望支撑具身 AI 模型的大规模训练，并迁移到真实世界。然而，尽管已有数十年进展，现有仿真仍只覆盖很窄的真实任务与物理现象。核心瓶颈在于：构建更好的仿真需要大量专家投入，因而无法规模化。为此，提出 Gym-Anything for Robotics 框架，由编码智能体（coding agent）创建逼真的机器人仿真。核心洞察是：仿真创建本质上是一项编码智能体任务，因此编码智能体的进步会直接转化为更丰富、更忠实的仿真。依据美国 GDP 数据挑选仿真任务，以覆盖有经济价值的物理工作；定义何为好的仿真与好的任务，并设计对抗式创建-审计循环（adversarial creation-audit loop），逐项按这些定义检查每个仿真与任务。基于该流水线构建 RoboWorld，包含 25 个仿真、25 个测试与 175 个训练 RL 环境，难度各异。即使最强的前沿模型在 RoboWorld 所有测试任务上均失败，平均奖励仅 0.06。基于 RoboWorld，提出多种提升前沿模型表现的新途径，包括跨多次尝试的上下文学习（in-context learning）、读取仿真器状态并在上下文中教导学生的特权教师（privileged teacher）、从模型自身特权经验中自辅助学习，以及生成跨多样场景演示以训练视觉-语言-动作模型（VLA，如 $\pi_{0.5}$）的代码策略。每种途径都带来显著提升。发布流水线、RL 环境与基准以促进后续研究。代码与数据见 https://gym-anything.github.io/robotics

关键要点

01现有机器人仿真仅覆盖窄范围任务与物理现象，瓶颈是依赖专家手工构建、无法规模化
02将仿真创建视为编码智能体任务，用对抗式创建-审计循环按 GDP 选任务并校验质量
03构建 RoboWorld：25 个仿真、25 个测试与 175 个训练 RL 环境，最强前沿模型平均奖励仅 0.06
04提出上下文学习、特权教师、自辅助学习与代码策略生成演示等提升途径，均显著改善表现
05开源流水线、RL 环境与基准，代码与数据发布于项目主页

解读

尚无解读。

原始英文摘要

arXiv:2610.40565v1 Announce Type: new Abstract: Realistic, fast robotics simulations could enable large-scale training of embodied AI models that transfer to the real world. However, despite decades of progress, available simulations cover only a narrow range of real-world tasks and physical phenomena. The core bottleneck is that creating better simulations requires significant expert effort and therefore does not scale. To address this, we introduce Gym-Anything for Robotics, a framework in which coding agents create realistic robotics simulations. Our core insight is that simulation creation is a coding agent task, so improvements in coding agents translate directly into richer and more faithful simulations. We select simulations based on U.S. GDP data to cover economically valuable physical work, define what makes a good simulation and a good task, and develop an adversarial creation-audit loop that checks each simulation and task against these definitions. Using this pipeline, we build RoboWorld, a set of 25 simulations with 25 test and 175 training RL environments of varying difficulty. Even the best frontier models fail on every RoboWorld test task, achieving an average reward of only 0.06. Building on RoboWorld, we introduce several new ways for frontier models to improve, including in-context learning across repeated attempts, a privileged teacher that reads the simulator state and teaches a student in context, self-assisted learning from the model's own privileged experience, and code policies that generate demonstrations across varied scenes to train vision-language-action models (VLAs) such as $\pi_{0.5}$. Each of these substantially improves performance. We release our pipeline, RL environments, and benchmark to facilitate future research. Code and data are available at https://gym-anything.github.io/robotics