arXiv 2610.10409 · cs.RO · 2026-10-07
RobotWorld: Benchmarking Multimodal Agents for Robot Use Across Diverse Tasks and Embodiments
Zhiqin Yang, Chenxin Li, Xiaomeng Hu, Yibin Liu, Weidong Huang, Jiankai Sun, Haitao Li, Zijian Wu 等 33 人
机构待核实
在最新榜 #226 本站收录于 2026-10-21
一句话要点
RobotWorld 用 84 个仿真任务（操作、移动、驾驶、飞行）检验通用多模态 Agent 能否操控机器人。
RobotWorld uses 84 simulated tasks (manipulation, locomotion, driving, aerial) to test multimodal agents at robot use.
阅读 arXiv 原文 PDF Hugging Face 讨论 代码仓库 项目页
主题与学科
同主题：机器人与具身
- Has GPT-6-Astra Achieved Embodied Intelligence? ▲465
- In-Context Learning for Robots: Methods and Applications ▲388
- GroundingPI: A Grounding Foundation Model towards Physical Intelligence with Visual Primitives ▲310
- Reflex: Visually Grounded Reactive Humanoid Control ▲154
- PanoVLN: Towards Effective Panoramic Vision-and-Language Navigation ▲151
数据说明
元数据来自 arXiv API（已核实）；点赞来自 Hugging Face 每日论文；引用来自 Semantic Scholar。最近刷新 2026-10-23。
要点由本站根据摘要自写，不转载摘要原文，请阅读原文。