# Measuring the Stability Assumption Behind Action Chunking

arXiv: https://arxiv.org/abs/2610.01626 · HTML: https://mlsanity.com/paper/2610.01626/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Imitation learning methods, one of 13 directions in Offline RL and imitation learning from data.
- Builds on 19 of its 22 references that are in the map.
- First posted on arXiv 1 Oct 2026.

## Placed in

- Offline RL and imitation learning from data / [Imitation learning methods](https://mlsanity.com/offlinerl/imitation-learning-methods/)

## Builds on (most cited first, 19 of 19)

- [Diffusion Policy: Visuomotor Policy Learning via Action Diffusion](https://mlsanity.com/paper/2303.04137/) (2023)
- [Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware](https://mlsanity.com/paper/2304.13705/) (2023)
- [GR00T N1: An Open Foundation Model for Generalist Humanoid Robots](https://mlsanity.com/paper/2503.14734/) (2025)
- [V-JEPA 2: Self-Supervised Video Models Enable Understanding, Prediction and Planning](https://mlsanity.com/paper/2506.09985/) (2025)
- [What Matters in Learning from Offline Human Demonstrations for Robot Manipulation](https://mlsanity.com/paper/2108.03298/) (2021)
- [robosuite: A Modular Simulation Framework and Benchmark for Robot Learning](https://mlsanity.com/paper/2009.12293/) (2020)
- [RoboCasa: Large-Scale Simulation of Everyday Tasks for Generalist Robots](https://mlsanity.com/paper/2406.02523/) (2024)
- [MimicGen: A Data Generation System for Scalable Robot Learning using Human Demonstrations](https://mlsanity.com/paper/2310.17596/) (2023)
- [DART: Noise Injection for Robust Imitation Learning](https://mlsanity.com/paper/1703.09327/) (2017)
- [Reinforcement Learning with Action Chunking](https://mlsanity.com/paper/2507.07969/) (2025)
- [Is Behavior Cloning All You Need? Understanding Horizon in Imitation Learning](https://mlsanity.com/paper/2407.15007/) (2024)
- [Learning Long-Context Diffusion Policies via Past-Token Prediction](https://mlsanity.com/paper/2505.09561/) (2025)
- [Adaptive Action Chunking at Inference-time for Vision-Language-Action Models](https://mlsanity.com/paper/2604.04161/) (2026)
- [The Pitfalls of Imitation Learning when Actions are Continuous](https://mlsanity.com/paper/2503.09722/) (2025)
- [Mixture of Horizons in Action Chunking](https://mlsanity.com/paper/2511.19433/) (2025)
- [Action Chunking and Exploratory Data Collection Yield Exponential Improvements in Behavior Cloning for Continuous Control](https://mlsanity.com/paper/2507.09061/) (2025)
- [Denoising Tells When to Replan: Denoising-Variance Adaptive Chunking for Flow-Based Robot Policies](https://mlsanity.com/paper/2606.03847/) (2026)
- [Why Does Action Chunking Improve Behavioral Cloning Performance in Robotic Control?](https://mlsanity.com/paper/2608.02547/) (2026)
- [Revisiting Open-Loop Execution in Robotics: Toward Reactive, Higher-Performing Policies](https://mlsanity.com/paper/2608.15938/) (2026)

## Abstract (arXiv)

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.