# DriveWLA：通过 World-Language-Action 建模统一未来动态与场景语义用于自动驾驶

> 原文链接：https://robotics.hellobike.com/papers/2026-10-13-2610-17903
> 频道：papers | 更新日期：2026-10-13 | 标签：arxiv, cs.RO
> 核心摘要：DriveWLA用MoT统一语义理解、未来动态与轨迹生成，并在开环、闭环及多基准零样本评估中验证。

# DriveWLA：通过 World-Language-Action 建模统一未来动态与场景语义用于自动驾驶

::: v-pre
> 原文链接：[DriveWLA: Unifying Future Dynamics and Scene Semantics through World-Language-Action Modeling for Autonomous Driving](https://arxiv.org/abs/2610.17903)（arXiv 2610.17903 · 发布日期：2026-10-13）

**DriveWLA: Unifying Future Dynamics and Scene Semantics through World-Language-Action Modeling for Autonomous Driving**

DriveWLA用MoT统一语义理解、未来动态与轨迹生成，并在开环、闭环及多基准零样本评估中验证。

## 核心要点

- VLMs与世界模型分别擅长语义理解和未来场景演化，但用于轨迹生成的有效融合仍具挑战。
- DriveWLA基于Mixture-of-Transformers，包含未来预测、语义理解和轨迹生成三个专家。
- Cross-Expert Attention控制专家间信息流并对齐语义与动态表示，Flow-GRPO后训练细化Action Expert。
- 实验显示其在开环和闭环评估中表现更优，并可零样本泛化到HUGSIM、nuScenes和Bench2Drive。

**验证范围与局限**：验证范围包括开环、闭环评估，以及HUGSIM、nuScenes、Bench2Drive上的零样本泛化。

## 英文摘要 (Abstract)

Vision-Language Models (VLMs) and world models have demonstrated strong capabilities in semantic understanding and future scene evolution for autonomous driving. However, effectively integrating both capabilities for trajectory generation remains challenging. In this work, we propose DriveWLA, a unified World-Language-Action model based on Mixture-of-Transformers (MoT) that jointly models semantic understanding, future scene dynamics, and trajectory generation. DriveWLA consists of three experts: a World Expert for future prediction, a Language Expert for semantic understanding, and an Action Expert for trajectory generation. We further introduce Cross-Expert Attention, which enables controlled information flow across experts while aligning semantic and dynamic representations. To improve trajectory generation beyond imitation learning, we apply Flow-GRPO reinforcement learning post-training to refine the Action Expert. Extensive experiments show that DriveWLA achieves superior performance in both open-loop and closed-loop evaluation. DriveWLA achieves 92.3 PDMS on NAVSIM v1 and 90.1 EPDMS on NAVSIM v2, demonstrating strong performance among VLM-based and world-model-based planners. Without specific fine-tuning, DriveWLA further demonstrates strong zero-shot generalization to HUGSIM, nuScenes, and Bench2Drive. These results validate the effectiveness of our unified World-Language-Action modeling, establishing DriveWLA as a generalizable framework for autonomous driving.

作者：Shuzheng Wang, Haoxuan Li, Shengzhen Li, Chenxi Wu, Yajing Sun, Lu Hou, Wenyu Liu, Xinggang Wang
:::

## 原始参考来源 (Sources)

- [arXiv 2610.17903：DriveWLA: Unifying Future Dynamics and Scene Semantics through World-Language-Action Modeling for Autonomous Driving](https://arxiv.org/abs/2610.17903)