动作链：用于机器人操作的轨迹自回归建模
Chain-of-Action: Trajectory Autoregressive Modeling for Robotic Manipulation
AI总结：
CoA通过逆向推理生成完整轨迹，结合连续动作表示、动态停止等设计，实现视觉-运动策略在机器人操作中的高效与灵活表现。
摘要：
我们提出了Chain-of-Action（CoA），一种基于轨迹自回归建模的新型视觉-运动策略范式。与传统方法预测下一步动作不同，CoA通过任务特定的目标进行显式逆向推理生成完整轨迹。这一过程在单一自回归结构中统一：（1）第一个标记对应一个稳定的键帧动作，编码任务特定的目标；（2）后续动作标记根据初始键帧和之前预测的动作自回归生成。这种逆向动作推理强制了全局到局部的结构，使每个局部动作紧密受最终目标约束。为了进一步实现动作推理结构，CoA结合了四种互补设计：连续动作标记表示；动态停止以生成可变长度轨迹；逆时间集成；以及多标记预测以平衡动作块建模与全局结构。结果表明，CoA在保持视觉-运动策略的灵活性和简洁性的同时，具有强大的空间泛化能力。实证上，我们观察到CoA在60个RLBench任务和8个现实世界操作任务上均实现了最先进的性能。
英文摘要：
We present Chain-of-Action (CoA), a novel visuo-motor policy paradigm built upon Trajectory Autoregressive Modeling. Unlike conventional approaches that predict next step action(s) forward, CoA generates an entire trajectory by explicit backward reasoning with task-specific goals through an action-level Chain-of-Thought (CoT) process. This process is unified within a single autoregressive structure: (1) the first token corresponds to a stable keyframe action that encodes the task-specific goals; and (2) subsequent action tokens are generated autoregressively, conditioned on the initial keyframe and previously predicted actions. This backward action reasoning enforces a global-to-local structure, allowing each local action to be tightly constrained by the final goal. To further realize the action reasoning structure, CoA incorporates four complementary designs: continuous action token representation; dynamic stopping for variable-length trajectory generation; reverse temporal ensemble; and multi-token prediction to balance action chunk modeling with global structure. As a result, CoA gives strong spatial generalization capabilities while preserving the flexibility and simplicity of a visuo-motor policy. Empirically, we observe CoA achieves the state-of-the-art performance across 60 RLBench tasks and 8 real-world manipulation tasks.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
