机器人学新招：世界模型与动作解耦，速度提升4.6倍
机器人操控通常需要同时预测未来画面和决定下一步动作，但现有方法把这两件事绑死在同一个时间节奏上，导致预测大量冗余的短期画面，浪费算力。这篇论文提出AHA-WAM，把任务拆成两个独立的AI：一个低频的“世界规划师”负责看长远场景变化，另一个高频的“动作执行师”只负责快速做动作。两个AI通过一种“观察引导的上下文路由”机制沟通，动作执行师可以随时从世界规划师那里获取长期背景，但不用等它重新计算。结果在模拟和真实任务中成功率最高达92.8%，控制频率达到24.17赫兹，比之前的方法快4.6倍。这不是你明天能用的技术，但它展示了机器人AI设计的一个新方向：让“想”和“动”各司其职，而不是混在一起。
📄 原文摘要(英文)
World-action models have emerged as a promising paradigm for robot manipulation, jointly modeling visual scene dynamics and actions to inject physical priors into policy learning. However, existing world-action models couple world prediction and action execution at the same temporal resolution, forcing the world branch to model near-term frame variations that are redundant and weakly informative. We posit that strictly binding world prediction and action execution to the same temporal rhythm may underutilize the potential of the video branch for embodied control. Therefore, we propose AHA-WAM, an Asynchronous Horizon-Adaptive World-Action Model built on a dual Diffusion Transformer (DiT) architecture that reorganizes world-action modeling around this temporal asymmetry. AHA-WAM instantiates the video DiT as a low-frequency world planner that maintains rolling key-value memory over past observations and exposes reusable layerwise latent context encoding long-horizon scene evolution, while a high-frequency action DiT executes short action chunks in closed loop by querying this context through layerwise joint attention. To support asynchronous execution, we introduce horizon-adaptive offset training and Observation-Guided Video-Context Routing (OVCR), which together let the action expert exploit long-horizon world context while remaining responsive to real-time execution state without rerunning the video DiT. Experiments on RoboTwin and real-world manipulation tasks show that AHA-WAM achieves state-of-the-art performance without any robot-data pretraining, attaining 92.80% average success on RoboTwin and 78.3% success across 4 real-world tasks, while reaching 24.17 Hz closed-loop control with a 4.59x speedup over Fast-WAM.