OmniBrain-0: A Unified Agentic Model for Embodied Intelligence
Computer Science > Computation and Language arXiv:2610.36013 (cs) [Submitted on 26 Oct 2026] Title: OmniBrain-0: A Unified Agentic Model for Embodied Intelligence. Authors: Xiaomi Embodied Intelligence Team: Yuechen Luo, Chong He, Fang Li, Shaoqing Xu, Jingkai Wang, Menglun Su, Jiaxin Liu, Shengyin Jiang, Zhiyi Lai, Long Chen, Longlong Wang, Liang Hu, Haochen Liu, Yuncheng Jiang, Hongwei Xie, Jieming Cui, Hanbing Li, Junjie Huang, Zhixiang Duan, Hexin Zhang, Weixiang Liang, Ying Gong, Yuhan Wang, Yong Pan, Pengna Li, Kangyi Wu, Ken Deng, Kun Ma, Fuxi Wen, Zhi-xin Yang, Hangjun Ye (University of Macau, Tsinghua University).
计算机科学 > 计算与语言 arXiv:2610.36013 (cs) [提交于 2026 年 10 月 26 日] 标题：OmniBrain-0：一种用于具身智能的统一代理模型。作者：小米具身智能团队：Yuechen Luo, Chong He, Fang Li, Shaoqing Xu, Jingkai Wang, Menglun Su, Jiaxin Liu, Shengyin Jiang, Zhiyi Lai, Long Chen, Longlong Wang, Liang Hu, Haochen Liu, Yuncheng Jiang, Hongwei Xie, Jieming Cui, Hanbing Li, Junjie Huang, Zhixiang Duan, Hexin Zhang, Weixiang Liang, Ying Gong, Yuhan Wang, Yong Pan, Pengna Li, Kangyi Wu, Ken Deng, Kun Ma, Fuxi Wen, Zhi-xin Yang, Hangjun Ye（澳门大学，清华大学）。
Abstract: General-purpose embodied intelligence requires robots to coordinate diverse skills over long horizons, track task progress, and recover from failures. Specialist policies alone do not ensure task-level reliability, while powerful multimodal agents rely on costly reasoning and external experience that compact models struggle to reproduce. We introduce OmniBrain-0, a lightweight unified embodied agent that coordinates manipulation and navigation through a shared high-level decision model, integrating task understanding, planning, progress tracking, and feedback-driven recovery.
摘要：通用具身智能要求机器人能够在长跨度内协调多种技能、跟踪任务进度并从故障中恢复。仅靠专业策略无法确保任务级的可靠性，而强大的多模态代理依赖于昂贵的推理和外部经验，这是紧凑型模型难以复现的。我们引入了 OmniBrain-0，这是一种轻量级的统一具身代理，它通过共享的高级决策模型来协调操作和导航，集成了任务理解、规划、进度跟踪和反馈驱动的恢复功能。
Our progressive training framework combines embodied pretraining, domain-specialized agent distillation, and multi-teacher on-policy distillation (MOPD). It distills powerful teachers’ interaction experience into manipulation and navigation experts, then consolidates their decision capabilities into OmniBrain-0. Unification therefore occurs within the learned decision model rather than only at the tool interface. We further introduce Rubric-Guided Recursive Self-Improvement (RSI), a data flywheel in which a Meta-Agent turns execution evidence into refined rubrics, step-level reminders, and verified trajectories for domain-expert updates before their reconsolidation into OmniBrain-0 through MOPD.
我们的渐进式训练框架结合了具身预训练、领域专业代理蒸馏和多教师在线策略蒸馏（MOPD）。它将强大教师的交互经验蒸馏到操作和导航专家中，然后将其决策能力整合到 OmniBrain-0 中。因此，统一化发生在学习到的决策模型内部，而不仅仅是在工具接口处。我们进一步引入了准则引导的递归自我改进（RSI），这是一个数据飞轮，其中元代理将执行证据转化为精炼的准则、步骤级提醒和经过验证的轨迹，用于领域专家更新，然后再通过 MOPD 将其重新整合到 OmniBrain-0 中。
We evaluate OmniBrain-0 on LIBERO-Pro, LIBERO-Plus, RoboCasa365, RoboTwin 2.0, and R2R-CE Val-Unseen, as well as multiple real-world manipulation tasks. OmniBrain-0 combines strong cross-domain task performance with substantially lower inference and execution overhead than its closed-source teacher, demonstrating efficient transfer and consolidation of embodied capabilities into a compact unified agent.
我们在 LIBERO-Pro、LIBERO-Plus、RoboCasa365、RoboTwin 2.0 和 R2R-CE Val-Unseen 以及多个现实世界操作任务上评估了 OmniBrain-0。OmniBrain-0 将强大的跨领域任务性能与比其闭源教师显著更低的推理和执行开销相结合，展示了将具身能力高效迁移并整合到一个紧凑的统一代理中的能力。