统一语言-动作理解与生成用于自动驾驶
Unifying Language-Action Understanding and Generation for Autonomous Driving
把语言与动作 token 统一进共享离散码本、并用轨迹生成描述文本的辅助理解目标，是可直接迁移到其他 VLA/具身策略学习中的对齐手段；C2F 两步粗到细解码则为自回归动作生成提供了可复用的加速范式。
原文依据 unifying language and action tokens into a shared discrete codebook, processed within a single multi-modal model
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论视觉-语言-动作（VLA）模型正成为端到端自动驾驶的一种有前景的范式，其价值在于能够利用世界知识并对复杂驾驶场景进行推理。然而，现有方法存在两个关键局限：语言指令与动作输出之间持续存在的错位，以及典型自回归动作生成固有的低效性。本文提出 LinkVLA，一种直接应对上述挑战以同时提升对齐性与效率的新架构。首先，我们通过将语言与动作 token 统一到共享的离散码本中，并在单一多模态模型内处理，从而建立结构层面的联系，从底层结构上强制实现跨模态一致性。其次，为建立深层语义联系，我们引入一个辅助的动作理解目标，训练模型由轨迹生成描述性文本，从而促进双向的语言-动作映射。最后，我们用两步粗到细生成方法（C2F）替代缓慢的逐步生成，该方法高效解码动作序列，节省 86% 的推理时间。在闭环驾驶基准上的实验表明，该方法在指令遵循准确率和驾驶性能上均取得稳定提升，同时降低了推理延迟。
英文原文摘要 1354 字符
Vision-Language-Action (VLA) models are emerging as a promising paradigm for end-to-end autonomous driving, valued for their potential to leverage world knowledge and reason about complex driving scenes. However, existing methods suffer from two critical limitations: a persistent misalignment between language instructions and action outputs, and the inherent inefficiency of typical auto-regressive action generation. In this paper, we introduce LinkVLA, a novel architecture that directly addresses these challenges to enhance both alignment and efficiency. First, we establish a structural link by unifying language and action tokens into a shared discrete codebook, processed within a single multi-modal model. This structurally enforces cross-modal consistency from the ground up. Second, to create a deep semantic link, we introduce an auxiliary action understanding objective that trains the model to generate descriptive captions from trajectories, fostering a bidirectional language-action mapping. Finally, we replace the slow, step-by-step generation with a two-step coarse-to-fine generation method (C2F) that efficiently decodes the action sequence, saving 86% inference time. Experiments on closed-loop driving benchmarks show consistent gains in instruction following accuracy and driving performance, alongside reduced inference latency.