外观
论文HF Daily Papers · arXiv 2610.353852 天前
EmbodiedLab评测具身智能Act-to-Know能力，最强模型原生控制成功率不足50%
本文由 AI 根据公开内容整理，可能有误，请以原文为准：EmbodiedLab: Benchmarking Embodied Agents That Act to Know（HF Daily Papers · arXiv 2610.35385）
核心内容
EmbodiedLab针对具身智能体在物理世界中通过交互探索未知规律（Act-to-Know）的能力提出新基准，填补了现有Manipulation与Scientific Discovery评测间的空白。该基准包含5个学科领域的30个定量科学实验及14种实验装置平台，要求机器人在参数动态重采样的物理仿真环境中自主设计流程、操作仪器并读取反馈。所有任务均经过可解性验证，评估指标同时涵盖知识获取准确性与物理状态达成度，代码已开源。
研究团队测试了6个前沿大模型及其配套Coding Agent，对比原生观察控制、增强感知规划及操作技能蒸馏三种配置。结果显示，即便最强模型在原生控制下成功率也不足50%，且各Agent的主要失败瓶颈均集中在与实验装置的物理交互环节。数据表明，对于能力较弱的模型，注入操作技能带来的性能增益显著高于单纯增加感知或运动规划模块，揭示了当前具身系统在精细操作层面的核心短板。
关键看点
- 评测范式：首创Act-to-Know基准，含5学科30实验14装置，参数动态采样防数据泄露
- 性能瓶颈：6大前沿模型原生控制成功率<50%，核心失败源于装置交互而非感知规划
- 技能迁移：操作技能蒸馏对弱模型提升幅度显著高于感知增强，验证操作先验价值
- 开源资产：完整Benchmark与评估代码已在GitHub开源，支持物理仿真环境复现
工程与产业观察
该评测揭示了具身智能从“执行指令”向“主动探索”跨越时的工程断层：当前SOTA模型的语义理解未能有效转化为精细物理交互能力。对于算法工程师，这提示在Sim-to-Real与通用策略训练中，应优先强化针对复杂器械的操作原语与力控反馈学习，而非仅堆叠视觉感知算力；产业侧则需关注高保真科学实验仿真环境的构建标准。
英文摘要 (Abstract)
Embodied agents are expected to accomplish real-world goals by deciding for themselves how and when to interact with their surroundings. For many such goals, changing the physical state of the world is the means rather than the end: the agent changes the state of a system through interaction, observes the feedback, and explores until it either achieves the goal itself (e.g., discovering physical laws) or learns what the goal requires; it acts to know. Manipulation benchmarks stop at acting to change, with the target state given; scientific discovery benchmarks evaluate knowing but replace physical experimentation with symbolic actions or numerical queries. We introduce EmbodiedLab, a benchmark in which a robot must act to know. Quantitative science experiments make this setting controllable: the parameters of the system are resampled for every instance, so the result cannot be known in advance, yet it can be scored exactly. Across 30 experiments in five disciplines on 14 apparatus platforms, the agent receives an objective and the apparatus, chooses its own procedure, and operates a physics-based simulated laboratory, reading the instruments through its cameras. Evaluation checks what the agent came to know, grounded in readings recorded during the episode, as well as the physical state the task requires; every experiment is verified to be solvable. We evaluate six frontier models in their developers' coding agents with native observation and control, with added perception and motion planning, and with manipulation skills distilled from the strongest agents. With native control, even the strongest agent succeeds less than half the time; for every agent, most failures arise in interacting with the apparatus, and manipulation skills help weaker agents more than perception and planning. The benchmark and its evaluation code are available at https://github.com/HexaFuture/EmbodiedLab.
Hugging Face Daily Papers 热度：9 upvotes 作者：Hongcheng Gao, Zelin Zheng, Alex Zhou, Jay Zhu, Shijia Ge, Yazhe Wang, Xiping Li, Jianshu Zeng, Di Wu, Xuan Shangguan, Yexing Wan, Lingyu He, Yongli Wang, Xiaohan Yu, Zhiqi Jia, Yue Liu, Bryan Hooi, Y…
来源
本页事实对应的公开页面，以原文为准。