SemanticVLA：通过协同显式轨迹与潜变量动作规划实现超越动作记忆的语义推理
SemanticVLA: Towards Semantic Reasoning over Action Memorization via Synergistic Explicit Trace and Latent Action Planning
对做具身智能与 VLA 的研究者，这篇给出了把 VLM 的预训练空间视觉定位能力直接转成可解释文本坐标轨迹（显式 trace）并与潜变量动作词元互补的规划范式，可作为替代纯动作监督微调的思路，也提供了指令改写鲁棒性的评测视角。
原文依据 Explicit trace reasoning generates interpretable spatial waypoints as textual coordinate sequences through the VLM's native language interface
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）模型已成为一种颇具前景的范式：预训练的视觉-语言模型（VLM）充当负责高层推理的系统 2，并与作为系统 1 的动作专家相连以完成低层运动控制。然而，现有工作未能真正发挥 VLM 的能力：VLM 产生的潜变量嵌入缺乏语义可解释性，对下游策略给出的引导模糊且不稳定；而仅使用动作监督又会使 VLM 退化为参数量庞大的融合编码器，只会记忆动作模式而无法进行泛化推理。为弥合这一差距，我们提出 SemanticVLA，通过协同的双路径设计发挥 VLM 的推理能力。显式轨迹推理借助 VLM 原生的语言接口生成可解释的空间路点，即以文本坐标序列的形式输出，直接复用其预训练的空间视觉定位能力，为任务规划提供思维过程。潜变量动作词元则通过学习以视觉观测为基础的紧凑视觉运动基元来补充轨迹推理，提供超越纯坐标预测的更细粒度动作表示。这种协同使轨迹推理能够利用 VLM 的多模态理解来精化潜变量词元预测，同时潜变量词元提供稳定且有视觉依据的引导，弥补轨迹在数值上的敏感性。SemanticVLA 在 LIBERO 上取得 97.0% 的平均成功率，在 SimplerEnv WidowX 上取得 65.1%，显著优于强基线。更重要的是，SemanticVLA 在两个仿真套件中于指令改写条件下均保持明显更稳定的性能，并在真实世界长程任务与推理密集型任务上展现出显著优势。通过语义显式的轨迹与视觉有据的潜变量动作词元连接 VLM 推理与动作专家，我们的方法实现了真正的推理而非动作记忆。
英文原文摘要 1958 字符
Vision-Language-Action (VLA) models have emerged as a promising paradigm where pretrained Vision-Language Models (VLMs) serve as System 2 for high-level reasoning, connected to action experts as System 1 for low-level motor control.However, current works fail to genuinely leverage VLM capabilities: VLMs produce latent embeddings that lack semantic interpretability, providing ambiguous and unstable guidance to downstream policies, while solely action supervision further causes VLMs to degenerate into parameter-heavy fusion encoders that memorize action patterns rather than perform generalized reasoning.To bridge this gap, we introduce SemanticVLA, which leverages VLM reasoning through synergistic dual-path design. Explicit trace reasoning generates interpretable spatial waypoints as textual coordinate sequences through the VLM's native language interface, directly reusing its pretrained spatial grounding to provide a thinking process for task planning. Latent action tokens complement trace reasoning by learning compact visuomotor primitives grounded in visual observations, providing more fine-grained action representations beyond pure coordinate prediction. This synergy enables trace reasoning to leverage VLM's multimodal understanding for refining latent token prediction, while latent tokens provide stable and grounded guidance that compensates for trace's numerical sensitivity.SemanticVLA achieves 97.0% average success rate on LIBERO and 65.1% on SimplerEnv WidowX, substantially outperforming strong baselines. More importantly, SemanticVLA maintains significantly more stable performance under instruction rephrasing in both simulation suites, and demonstrates strong advantages on real-world long-horizon and reasoning-intensive tasks.By bridging VLM reasoning and action expert through semantically explicit trace and visually grounded latent action tokens, our approach enables genuine reasoning rather than action memorization.