# Temporal Action Selection for Action Chunking

arXiv: https://arxiv.org/abs/2511.04421 · HTML: https://mlsanity.com/paper/2511.04421/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Placed in Robot manipulation imitation learning, one of 13 directions in Offline RL and imitation learning from data.
- First posted on arXiv 6 Nov 2025.

## Placed in

- Offline RL and imitation learning from data / [Robot manipulation imitation learning](https://mlsanity.com/offlinerl/robot-manipulation-imitation-learning/)

## Built on by (most cited first, 4 of 4)

- [Median Temporal Ensembling: Training-Free Robust Aggregation for Action-Chunked Visuomotor Policies](https://mlsanity.com/paper/2609.27167/) (2026)
- [Dynamic Manipulation with World-Action Models via Counterfactual Planning](https://mlsanity.com/paper/2609.33172/) (2026)
- [Adaptive Mean Flow for Responsive Closed-Loop Robot Control](https://mlsanity.com/paper/2610.06089/) (2026)
- [DexNinja: Learning Robust Dexterous Cutting Policy with a Real-to-Sim-to-Real Data Engine](https://mlsanity.com/paper/2610.17208/) (2026)

## Abstract (arXiv)

Action chunking is a widely adopted approach in Learning from Demonstration (LfD). By modeling multi-step action chunks rather than single-step actions, action chunking significantly enhances modeling capabilities for human expert policies. However, because action chunking makes a single decision only after a complete action block has been executed, the resulting reduction in decision frequency restricts the utilization of real-time observations, impairing reactivity in dynamic or noisy environments. Existing efforts to address this issue have primarily resorted to trading off reactivity against decision consistency, without achieving both. To address this limitation, we propose a novel algorithm, Temporal Action Selection (TAS), which caches predicted action chunks from multiple timesteps and dynamically selects the optimal action through a lightweight selector network. TAS achieves balanced optimization across both reactivity and decision consistency. Experiments across multiple tasks with diverse base policy architectures show that TAS significantly improves success rates. Furthermore, integrating TAS as a base policy with residual reinforcement learning (RL) improves both training efficiency and the performance ceiling. Experiments in both simulation and physical robots confirm the method's efficacy.