Saved in:
Bibliographic Details
Main Authors: Huang, Minqing, Xiang, Yujiao, Liang, Zihan, Huang, Jiajie, Wang, Jingqi, Xu, Zhi, Tan, Feiyang, Zhou, Hangning, Yang, Mu, Che, Gong
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2605.10426
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913121972518912
author Huang, Minqing
Xiang, Yujiao
Liang, Zihan
Huang, Jiajie
Wang, Jingqi
Xu, Zhi
Tan, Feiyang
Zhou, Hangning
Yang, Mu
Che, Gong
author_facet Huang, Minqing
Xiang, Yujiao
Liang, Zihan
Huang, Jiajie
Wang, Jingqi
Xu, Zhi
Tan, Feiyang
Zhou, Hangning
Yang, Mu
Che, Gong
contents Vision-Language-Action (VLA) models have emerged as a promising paradigm for end-to-end autonomous driving. However, existing reasoning mechanisms still struggle to provide planning-oriented intermediate representations: textual Chain-of-Thought (CoT) fails to preserve continuous spatiotemporal structure, while latent world reasoning remains difficult to use as a direct condition for action generation. In this paper, we propose CoWorld-VLA, a multi-expert world reasoning framework for autonomous driving, where world representations serve as explicit conditions to guide action planning. CoWorld-VLA extracts complementary world information through multi-source supervision and encodes it into expert tokens within the VLA, thereby providing planner-accessible conditioning signals. Specifically, we construct four types of tokens: semantic interaction, geometric structure, dynamic evolution, and ego trajectory tokens, which respectively model interaction intent, spatial structure, future temporal dynamics, and behavioral goals. During action generation, CoWorld-VLA employs a diffusion-based hierarchical multi-expert fusion planner, which is coupled with scene context throughout the joint denoising process to generate continuous ego trajectories. Experiments show that CoWorld-VLA achieves competitive results in both future scene generation and planning on the NAVSIM v1 benchmark, demonstrating strong performance in collision avoidance and trajectory accuracy. Ablation studies further validate the complementarity of expert tokens and their effectiveness as planning conditions for action generation. Code will be available at https://github.com/AFARI-Research/CoWorld-VLA.
format Preprint
id arxiv_https___arxiv_org_abs_2605_10426
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle CoWorld-VLA: Thinking in a Multi-Expert World Model for Autonomous Driving
Huang, Minqing
Xiang, Yujiao
Liang, Zihan
Huang, Jiajie
Wang, Jingqi
Xu, Zhi
Tan, Feiyang
Zhou, Hangning
Yang, Mu
Che, Gong
Computer Vision and Pattern Recognition
Artificial Intelligence
Vision-Language-Action (VLA) models have emerged as a promising paradigm for end-to-end autonomous driving. However, existing reasoning mechanisms still struggle to provide planning-oriented intermediate representations: textual Chain-of-Thought (CoT) fails to preserve continuous spatiotemporal structure, while latent world reasoning remains difficult to use as a direct condition for action generation. In this paper, we propose CoWorld-VLA, a multi-expert world reasoning framework for autonomous driving, where world representations serve as explicit conditions to guide action planning. CoWorld-VLA extracts complementary world information through multi-source supervision and encodes it into expert tokens within the VLA, thereby providing planner-accessible conditioning signals. Specifically, we construct four types of tokens: semantic interaction, geometric structure, dynamic evolution, and ego trajectory tokens, which respectively model interaction intent, spatial structure, future temporal dynamics, and behavioral goals. During action generation, CoWorld-VLA employs a diffusion-based hierarchical multi-expert fusion planner, which is coupled with scene context throughout the joint denoising process to generate continuous ego trajectories. Experiments show that CoWorld-VLA achieves competitive results in both future scene generation and planning on the NAVSIM v1 benchmark, demonstrating strong performance in collision avoidance and trajectory accuracy. Ablation studies further validate the complementarity of expert tokens and their effectiveness as planning conditions for action generation. Code will be available at https://github.com/AFARI-Research/CoWorld-VLA.
title CoWorld-VLA: Thinking in a Multi-Expert World Model for Autonomous Driving
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2605.10426