Enregistré dans:
| Auteurs principaux: | , , , , , , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | https://arxiv.org/abs/2602.14979 |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866917276819652608 |
|---|---|
| author | Dang, Ronghao Guo, Jiayan Hou, Bohan Leng, Sicong Li, Kehan Li, Xin Liu, Jiangpin Mao, Yunxuan Wang, Zhikai Yuan, Yuqian Zhu, Minghao Lin, Xiao Bai, Yang Jiang, Qian Zhao, Yaxi Zeng, Minghua Gao, Junlong Jiang, Yuming Cen, Jun Huang, Siteng Wang, Liuyi Zhang, Wenqiao Liu, Chengju Yang, Jianfei Lu, Shijian Zhao, Deli |
| author_facet | Dang, Ronghao Guo, Jiayan Hou, Bohan Leng, Sicong Li, Kehan Li, Xin Liu, Jiangpin Mao, Yunxuan Wang, Zhikai Yuan, Yuqian Zhu, Minghao Lin, Xiao Bai, Yang Jiang, Qian Zhao, Yaxi Zeng, Minghua Gao, Junlong Jiang, Yuming Cen, Jun Huang, Siteng Wang, Liuyi Zhang, Wenqiao Liu, Chengju Yang, Jianfei Lu, Shijian Zhao, Deli |
| contents | Despite rapid progress in multimodal foundation models, embodied intelligence community still lacks a unified, physically grounded foundation model that integrates perception, reasoning, and planning within real-world spatial-temporal dynamics. We introduce RynnBrain, an open-source spatiotemporal foundation model for embodied intelligence. RynnBrain strengthens four core capabilities in a unified framework: comprehensive egocentric understanding, diverse spatiotemporal localization, physically grounded reasoning, and physics-aware planning. The RynnBrain family comprises three foundation model scales (2B, 8B, and 30B-A3B MoE) and four post-trained variants tailored for downstream embodied tasks (i.e., RynnBrain-Nav, RynnBrain-Plan, and RynnBrain-VLA) or complex spatial reasoning tasks (i.e., RynnBrain-CoP). In terms of extensive evaluations on 20 embodied benchmarks and 8 general vision understanding benchmarks, our RynnBrain foundation models largely outperform existing embodied foundation models by a significant margin. The post-trained model suite further substantiates two key potentials of the RynnBrain foundation model: (i) enabling physically grounded reasoning and planning, and (ii) serving as a strong pretrained backbone that can be efficiently adapted to diverse embodied tasks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_14979 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | RynnBrain: Open Embodied Foundation Models Dang, Ronghao Guo, Jiayan Hou, Bohan Leng, Sicong Li, Kehan Li, Xin Liu, Jiangpin Mao, Yunxuan Wang, Zhikai Yuan, Yuqian Zhu, Minghao Lin, Xiao Bai, Yang Jiang, Qian Zhao, Yaxi Zeng, Minghua Gao, Junlong Jiang, Yuming Cen, Jun Huang, Siteng Wang, Liuyi Zhang, Wenqiao Liu, Chengju Yang, Jianfei Lu, Shijian Zhao, Deli Robotics Despite rapid progress in multimodal foundation models, embodied intelligence community still lacks a unified, physically grounded foundation model that integrates perception, reasoning, and planning within real-world spatial-temporal dynamics. We introduce RynnBrain, an open-source spatiotemporal foundation model for embodied intelligence. RynnBrain strengthens four core capabilities in a unified framework: comprehensive egocentric understanding, diverse spatiotemporal localization, physically grounded reasoning, and physics-aware planning. The RynnBrain family comprises three foundation model scales (2B, 8B, and 30B-A3B MoE) and four post-trained variants tailored for downstream embodied tasks (i.e., RynnBrain-Nav, RynnBrain-Plan, and RynnBrain-VLA) or complex spatial reasoning tasks (i.e., RynnBrain-CoP). In terms of extensive evaluations on 20 embodied benchmarks and 8 general vision understanding benchmarks, our RynnBrain foundation models largely outperform existing embodied foundation models by a significant margin. The post-trained model suite further substantiates two key potentials of the RynnBrain foundation model: (i) enabling physically grounded reasoning and planning, and (ii) serving as a strong pretrained backbone that can be efficiently adapted to diverse embodied tasks. |
| title | RynnBrain: Open Embodied Foundation Models |
| topic | Robotics |
| url | https://arxiv.org/abs/2602.14979 |