Why Does Action Chunking Improve Behavioral Cloning Performance in Robotic Control?
2026-08-03 • Robotics
Robotics
AI summaryⓘ
The authors studied why action chunking, which means planning multiple actions at once, helps robots perform better. They found that common explanations like smoother actions or simpler decision-making don't fully explain its success. Instead, they showed that action chunking works partly because it can consider past information better and reduce errors, and also because it acts like a group of different models working together (an ensemble), making it more reliable. They then designed new methods that copy and improve on these ensemble benefits to get even better robot control without always having to chunk actions.
action chunkingrobotic control policiesMarkovian policynon-Markovian expressivitycompounding errordelayed policiesmodel ensembletemporal relationshipsrobustnessgeneralization
Authors
Filippo Lazzati, Kyle Stachowicz, William Chen, Alberto Maria Metelli, Andrew Wagenmaker, Sergey Levine
Abstract
Action chunking---predicting and executing multiple actions instead of a single action---has proven to be a critical component for learning effective robotic control policies. However, our precise understanding of why action chunking improves performance has remained limited. In this work we seek to close this gap. Through rigorous experimental evaluations in both simulated and real-world settings, we show that existing hypotheses for the success of action chunking---temporal consistency, horizon reduction, and representation learning---fail to explain the success of action chunking. Instead, we find that action chunking benefits from greater non-Markovian expressivity and reduced compounding error compared to Markovian policies, but, in many settings of interest, these effects can be fully captured by delayed policies, which at each step predict a single action based on the observation $k$ steps in the past. We then show that there exists an additional benefit of action chunking that we refer to as implicit ensembling. In particular, by learning a diversity of temporal relationships (that is, $a_t | o_t, a_t | o_{t-1}, \ldots$), action-chunked policies exhibit behavior matching that of a model ensemble, increasing their robustness and generalization ability over policies that only learn a single temporal relationship. Building on these insights, we show that in simulated and real-world robotic control settings, we can match the performance of action chunking without action chunking---by deploying an action chunking policy as an ensemble of policies with randomized delays. Furthermore, we propose a policy class that amplifies the benefits of action chunking by explicitly instantiating an ensemble, and which we show significantly improves over the performance of action chunking in many domains.