Motus2 README

The project's README from github.com/shengshu-ai/Motus2. As one markdown file: README.md.

Motus2: A Self-Evolving General World Model for Dexterous Manipulation

Project Page arXiv Hugging Face

Overview

General embodied agents should perceive, predict, act, evaluate, and improve within a unified system. World models have shown great promise in building such agents, yet existing models typically append an action output head to a world simulator, without coupling them into a closed decision-and-learning loop for policy improvement. We present Motus2, a self-evolving general world model for dexterous manipulation. Motus2 advances world modeling through model scaling and data scaling. For model scaling, a single model with shared weights exposes three control interfaces: a policy (world–action model), a simulator (action-conditioned world model), and an evaluator (value model). The policy proposes candidate action chunks, the simulator predicts their visual consequences, and the evaluator assesses the predicted outcomes. Their coupling forms a closed decision-and-learning loop for policy improvement. This formulation uses curated expert demonstrations for action learning, while failed and suboptimal interactions provide valuable evidence for dynamics modeling and value learning. For data scaling, Motus2 progresses from large-scale monocular egocentric data to synchronized stereo egocentric data, followed by robot-domain adaptation with robot trajectories and supplementary human–robot alignment data. Motus2 further studies global-autoregressive and hybrid-memory extensions of its sliding-window context, adds tactile feedback for contact-aware control, and is instantiated on a fully biomimetic platform with stereo vision, dual arms, dual dexterous hands, and tactile sensing. Together, egocentric data scaling and closed-loop general world model scaling provide a general path toward self-evolving dexterous manipulation.

Motus2 overview

Updates

Open-Source Roadmap

Citation

If you find Motus2 useful in your research, please cite:

@misc{bi2026motus2,
  title         = {Motus2: A Self-Evolving General World Model for Dexterous Manipulation},
  author        = {Hongzhe Bi and Zihao Zhou and Yihang Tang and Jingrui Pang and
                   Shuhe Huang and Haitian Liu and Runqing Wang and Shuai Huang and
                   Yichen Wang and Yiming Cheng and Ruowen Zhao and Zhenghua Li and
                   Hengkai Tan and Xiaolong Liu and Jinhui Wan and Jiabao Liu and
                   Min Zhao and Fan Bao and Jun Zhu},
  year          = {2026},
  eprint        = {2608.30237},
  archivePrefix = {arXiv},
  primaryClass  = {cs.RO},
  url           = {https://arxiv.org/abs/2608.30237}
}

Motus2 (shengshu-ai/Motus2)

Searching Database…

685 ·GitHub
Motus2
这是什么？Motus2 是一个面向灵巧机器人操作的自进化通用世界模型。一个共享权重的单一模型同时充当策略、模拟器和评估器，并把三者耦合为闭环的决策与学习回路，使系统能够感知、预测、行动并自我改进。
这是开箱即用的模型，还是更像一个研究框架？它是研究性质的模型/代码库，不是拿来即用的工具：单一模型暴露出三个协同工作的控制接口（策略、模拟器、评估器），设计上要靠大规模第一视角和机器人数据来扩展，而不是开箱即用。
我怎么让它跑起来？目前还不行——尚未发布任何代码或检查点。作者计划在 2026 年 9 月内逐步发布预训练检查点和训练代码（视频预训练、价值/MBRL 训练、记忆基础设施、触觉代码）。
现在有什么要注意的？开源路线图上的所有条目都还未勾选——预训练检查点、视频预训练代码、价值训练代码、MBRL 代码、记忆基础设施和触觉代码都仍待发布。
策略、模拟器和评估器接口具体是做什么的？策略提出候选动作块，模拟器预测其视觉后果，评估器评估这些预测结果的好坏，从而形成用于改进策略的闭环决策与学习回路。
Motus2 如何扩展它的训练数据？它从大规模单目第一视角视频进阶到同步双目第一视角数据，再适配到机器人轨迹，并辅以人机对齐数据来完成域迁移。
它为什么被称为“自进化”？因为失败和次优的交互（而不仅是精选的专家演示）也会反馈到动力学和价值学习中，让模型通过耦合的模拟器-评估器回路改进自己的策略，而不是只依赖人工标注的成功数据。
Motus2 面向什么硬件？一个具备双目视觉、双臂、双灵巧手和触觉感知的全仿生机器人平台，并加入触觉反馈以实现接触感知控制。
What are similar projects?Zero-WAM, lerobot, WholebodyVLA, Zetta-Embodiment, Embodied-AI-Guide, flex-pi, RoboICL, video-prediction-policy-2
Which topics is it in?physical-ai, generative-model, self-evolving-agents, pretraining, deep-reinforcement-learning, robot, world-models, embodied-ai, pretrained-models, paper, robotics, simulation
Creator
shengshu-ai
Size
<0.1 MB
Stars
685
Updated
CommitSep 11
Age
1mo
Commits · 90 Days