Measuring the Stability Assumption Behind Action Chunking

📄 arXiv: 2610.01626v1 📥 PDF

作者: Aryan Goyal

分类: cs.AI

发布日期: 2026-10-01

备注: 18 pages, 9 figures, 18 tables

💡 一句话要点

研究动作分块背后的稳定性假设以提升行为克隆性能

🎯 匹配领域: 支柱一：机器人控制 (Robot Control) 支柱二：RL算法与架构 (RL & Architecture)

关键词: 动作分块 行为克隆 误差传播 开放环路 闭合环路 机器人控制 自适应策略

📋 核心要点

  1. 现有的动作分块机制未能充分解释动作误差在系统中的传播特性，导致对行为克隆性能提升的理解不足。
  2. 本文通过研究动作误差在开放环路和闭合环路下的传播特性，提出了一种新的评估方法，关注误差的增长和收缩速率。
  3. 在十二个操作任务中，发现稳定状态稀少，误差放大普遍存在，且传播速率受拟合时间窗口的显著影响。

📝 摘要（中文）

动作分块通过行为克隆提高了策略的性能，已有多种机制被提出以解释其原因。本文研究了动作误差在系统中的变化，分别在开放环路和闭合环路下注入小的动作误差，测量其增长或缩小的速度。研究发现，稳定状态较少，而误差放大现象普遍存在。此外，测量的传播速率与拟合时间窗口密切相关，短时间窗口可能高估长期传播。最后，训练预测器发现，开放环路状态可以通过视觉和本体感知恢复，而闭合环路传播则部分可恢复，表明现有的误差累积理论并不足以解释动作分块的全部现象。

🔬 方法详解

问题定义：本文旨在解决动作分块中动作误差传播的特性，现有方法未能充分解释误差在系统中的变化，尤其是在不同执行模式下的表现。

核心思路：通过在每个状态注入小的动作误差，测量其在开放环路和闭合环路下的传播速率，评估状态的稳定性，进而揭示动作分块的内在机制。

技术框架：研究分为两个主要阶段：第一阶段是注入动作误差并记录其变化，第二阶段是分析不同状态下的传播速率，标记为收缩、扩张或未解决。

关键创新：本研究的创新在于通过直接测量动作误差的传播特性，挑战了传统的误差累积理论，表明开放环路和闭合环路下的动态行为并不一致。

关键设计：在实验中，采用了不同的拟合时间窗口来评估传播速率，发现短时间窗口可能导致对长期传播的高估，同时设计了预测器以从视觉和本体感知中恢复状态信息。

🖼️ 关键图片

fig_0
img_1
img_2

📊 实验亮点

实验结果显示，在十二个操作任务中，稳定状态的出现频率极低，而误差放大现象普遍存在。通过训练的预测器能够从视觉和本体感知中恢复开放环路状态，闭合环路传播的恢复率则相对较低，表明现有的学习方法在应对误差时的局限性。

🎯 应用场景

该研究的潜在应用领域包括机器人控制、自动驾驶和人机交互等，能够帮助设计更鲁棒的控制策略，提高系统在面对不确定性时的反应能力，未来可能对智能系统的自适应能力产生深远影响。

📄 摘要（原文）

Action chunking improves the performance of policies learned by behavioural cloning, and several mechanisms have been proposed to explain why, including temporal consistency, horizon reduction, representation learning, and reduced error compounding. We instead study what happens to an action error once it enters the system. At each state, we inject a small action error and measure how fast it grows or shrinks under two execution regimes: open-loop, where the rest of the chunk is replayed without replanning, and closed-loop, where the policy replans after the perturbation. The fitted rate labels each state as contracting, expanding, or unresolved. Across twelve manipulation tasks from three benchmark suites, we find that confidently stable states are rare, while error amplification is common among states whose propagation rate can be resolved. We further find that the measured propagation rate depends strongly on the fitting horizon: amplification is typically front-loaded, so short windows can overestimate longer-horizon propagation. Finally, we train predictors on these labels and find that a state's open-loop regime can be recovered from camera frames and proprioception alone, while its closed-loop propagation is only partially recoverable because it also depends on how the policy acts after the perturbation. These results suggest that error-compounding arguments alone do not provide a complete account of action chunking: neither passive open-loop dynamics nor policy replanning consistently contracts an injected error, and replanning rarely turns open-loop amplification into confident contraction. This suggests that closed-loop reactivity should be trained explicitly, using perturbation- and tree-coverage-oriented training to expose policies to deviations they must recover from, rather than expected to emerge reliably from standard imitation learning.