# Policy learning in SE(3) action spaces

arXiv: https://arxiv.org/abs/2010.02798 · HTML: https://mlsanity.com/paper/2010.02798/

## Key facts

- Cited by 14 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 6 Oct 2020.

## Built on by (most cited first, 14 of 14)

- [Invariant Slot Attention: Object Discovery with Slot-Centric Reference Frames](https://mlsanity.com/paper/2302.04973/) (2023)
- [SEIL: Simulation-augmented Equivariant Imitation Learning](https://mlsanity.com/paper/2211.00194/) (2022)
- [Leveraging Symmetries in Pick and Place](https://mlsanity.com/paper/2308.07948/) (2023)
- [On Robot Grasp Learning Using Equivariant Models](https://mlsanity.com/paper/2306.06489/) (2023)
- [Action Priors for Large Action Spaces in Robotics](https://mlsanity.com/paper/2101.04178/) (2021)
- [Factored World Models for Zero-Shot Generalization in Robotic Manipulation](https://mlsanity.com/paper/2202.05333/) (2022)
- [QDP: Learning to Sequentially Optimise Quasi-Static and Dynamic Manipulation Primitives for Robotic Cloth Manipulation](https://mlsanity.com/paper/2303.13320/) (2023)
- [Learning from Pixels with Expert Observations](https://mlsanity.com/paper/2306.13872/) (2023)
- [Coarse-to-Fine 3D Keyframe Transporter](https://mlsanity.com/paper/2502.01773/) (2025)
- [Towards Autonomous Reinforcement Learning for Real-World Robotic Manipulation with Large Language Models](https://mlsanity.com/paper/2503.04280/) (2025)
- [Generalizable Hierarchical Skill Learning via Object-Centric Representation](https://mlsanity.com/paper/2510.21121/) (2025)
- [Symmetries-enhanced Multi-Agent Reinforcement Learning](https://mlsanity.com/paper/2501.01136/) (2025)
- [Visual Foresight With a Local Dynamics Model](https://mlsanity.com/paper/2206.14802/) (2022)
- [Pix2Act: Image-Space Manipulation Policies with Equivariant Augmentation](https://mlsanity.com/paper/2607.11167/) (2026)

## Abstract (arXiv)

In the spatial action representation, the action space spans the space of target poses for robot motion commands, i.e. SE(2) or SE(3). This approach has been used to solve challenging robotic manipulation problems and shows promise. However, the method is often limited to a three dimensional action space and short horizon tasks. This paper proposes ASRSE3, a new method for handling higher dimensional spatial action spaces that transforms an original MDP with high dimensional action space into a new MDP with reduced action space and augmented state space. We also propose SDQfD, a variation of DQfD designed for large action spaces. ASRSE3 and SDQfD are evaluated in the context of a set of challenging block construction tasks. We show that both methods outperform standard baselines and can be used in practice on real robotics systems.