Minghao Zhu, Bohan Hou, Kun Xiang, Zixuan Wang, Yinzhou Tang, Yifei Yang, Zhikai Wang, Tianyi Zhang, Kehan Li, Haoyu Zhao, Siteng Huang, Leigang Qu, Sicong Leng, Dongchi Huang, Mingxiu Chen, Shijian Lu, Yue Wang, Xiaodan Liang, Zhenyu Zeng, Deli Zhao, Chengju Liu, Jianfei Yang, Xin Li
RynnUnify is a 16B unified embodied model that integrates visual planning, physical reasoning, and long-horizon anticipation through interleaved video-language reasoning. It enables agentic decision-making and robot control via reinforcement learning and world-action extensions, outperforming prior models on embodied understanding and task success.
This work advances the development of a unified cognitive architecture for embodied agents, bridging perception, reasoning, and action in real-world environments. Its success in complex robotic tasks demonstrates practical pathways toward autonomous, adaptive robots capable of long-horizon planning.
Thank you to arXiv for use of its open access interoperability. arxiv.org. aibrief.ai is not affiliated with or endorsed by arXiv.