AtomicVLA：释放机器人中原子技能学习的潜力
AtomicVLA: Unlocking the Potential of Atomic Skill Learning in Robots
对做视觉-语言-动作模型的人，这篇给出一种把长程操作拆成原子技能、用技能引导的混合专家（SG-MoE）加路由编码器动态分配专家的做法，可直接借鉴到需要持续学习新技能的机器人策略中，替代单一动作解码器。
原文依据 we propose AtomicVLA, a unified planning-and-execution framework that jointly generates task-level plans, atomic skill abstractions, and fine-grained actions
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论近期视觉-语言-动作（VLA）模型的进展在机器人操作任务上展现出可观潜力。然而，真实世界的机器人任务往往涉及长程、多步的问题求解，并需要面向持续技能获取的泛化能力，其范围超出单一动作或技能。这些挑战为现有 VLA 模型带来了显著障碍：它们使用在聚合数据上训练的单体式动作解码器，导致可扩展性较差。为应对这些挑战，我们提出 AtomicVLA，一个统一规划与执行的框架，能够联合生成任务级规划、原子技能抽象与细粒度动作。AtomicVLA 通过技能引导的混合专家（SG-MoE）构建可扩展的原子技能库，其中每个专家专注于掌握通用而精确的原子技能。此外，我们引入一种灵活的路由编码器，可自动为新技能分配专属的原子专家，从而实现持续学习。我们通过大量实验验证了所提方法。在仿真中，AtomicVLA 在 LIBERO 上以 2.4% 的优势超过 π_0，在 LIBERO-LONG 上超过 10%，并在 CALVIN 上以 0.22 和 0.25 的平均任务长度优势超过 π_0 与 π_0.5。此外，我们的 AtomicVLA 在真实世界长程任务与持续学习上分别以 18.3% 和 21% 的幅度持续超越基线。这些结果凸显了原子技能抽象与动态专家组合在长程与终身机器人任务中的有效性。
英文原文摘要 1500 字符
Recent advances in Visual-Language-Action (VLA) models have shown promising potential for robotic manipulation tasks.However, real-world robotic tasks often involve long-horizon, multi-step problem-solving and require generalization for continual skill acquisition, extending beyond single actions or skills. These challenges present significant barriers for existing VLA models, which use monolithic action decoders trained on aggregated data, resulting in poor scalability.To address these challenges, we propose AtomicVLA, a unified planning-and-execution framework that jointly generates task-level plans, atomic skill abstractions, and fine-grained actions. AtomicVLA constructs a scalable atomic skill library through a Skill-Guided Mixture-of-Experts (SG-MoE), where each expert specializes in mastering generic yet precise atomic skills. Furthermore, we introduce a flexible routing encoder that automatically assigns dedicated atomic experts to new skills, enabling continual learning.We validate our approach through extensive experiments. In simulation, AtomicVLA outperforms \pi_ 0 by 2.4% on LIBERO, 10% on LIBERO-LONG, and outperforms \pi_ 0 and \pi_ 0.5 by 0.22 and 0.25 in average task length on CALVIN. Additionally, our AtomicVLA consistently surpasses baselines by 18.3% and 21% in real-world long-horizon tasks and continual learning. These results highlight the effectiveness of atomic skill abstraction and dynamic expert composition for long-horizon and lifelong robotic tasks.