ZhiYuan Robotics launches ACoT-VLA and open sources it
The ACoT-VLA (Action Chain-of-Thought) architecture proposed by Zhiyuan Robotics (AGIBOT) in collaboration with Beihang University has officially been selected for the top conference in computer vision, CVPR 2026. This research breaks the traditional bottleneck of VLA models, which are "semantically strong but action weak," and for the first time proposes a chain of thought paradigm for reasoning in the "action space." Currently, ACoT-VLA has been officially open-sourced as the baseline model for the AGIBOT WORLD CHALLENGE, assisting global developers in tackling complex robotic manipulation challenges
Source: Wallstreetcn The copyright of this article belongs to the original author/organization.
The views expressed herein are solely those of the author and do not reflect the stance of the platform. The content is intended for investment reference purposes only and shall not be considered as investment advice. Please contact us if you have any questions or suggestions regarding the content services provided by the platform.
Post your comment
No Comments