mlsanityData to 7 Oct 2026

Equivariant Reinforcement Learning under Partial Observability

arXiv 2408.14336 ↗First posted 26 Aug 2024cs.RO cs.AI cs.CVcited by 17

Not on the map. Why?

Abstract (arXiv)

Incorporating inductive biases is a promising approach for tackling challenging robot learning domains with sample-efficient solutions. This paper identifies partially observable domains where symmetries can be a useful inductive bias for efficient learning. Specifically, by encoding the equivariance regarding specific group symmetries into the neural networks, our actor-critic reinforcement learning agents can reuse solutions in the past for related scenarios. Consequently, our equivariant agents outperform non-equivariant approaches significantly in terms of sample efficiency and final performance, demonstrated through experiments on a range of robotic tasks in simulation and real hardware.

Key facts. Cited by 17 papers in the map (data through 7 Oct 2026). Builds on 47 of its 65 references that are in the map. First posted on arXiv 26 Aug 2024.

Builds on

47 of 65 references.

References · 47

  1. 2017Proximal Policy Optimization Algorithmscited by 14,441
  2. 2018Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actorcited by 4,635
  3. 2016Asynchronous Methods for Deep Reinforcement Learningcited by 2,218
  4. 2018World Modelscited by 1,444
  5. 2015Convolutional LSTM Network: A Machine Learning Approach for Precipitation Nowcastingcited by 1,428
  6. 2023Mastering Diverse Domains through World Modelscited by 1,193
  7. 2016Group Equivariant Convolutional Networkscited by 1,154
  8. 2020Mastering Atari with Discrete World Modelscited by 990
  9. 2021E(n) Equivariant Graph Neural Networkscited by 951
  10. 2018DeepMind Control Suitecited by 935
  11. 2020CURL: Contrastive Unsupervised Representations for Reinforcement Learningcited by 745
  12. 2018Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point cloudscited by 735
35 more
  1. 2021E(3)-Equivariant Graph Neural Networks for Data-Efficient and Accurate Interatomic Potentialscited by 701
  2. 2020Image Augmentation Is All You Need: Regularizing Deep Reinforcement Learning from Pixelscited by 573
  3. 2015Deep Recurrent Q-Learning for Partially Observable MDPscited by 457
  4. 2020Reinforcement Learning with Augmented Datacited by 452
  5. 2019General E(2)-Equivariant Steerable CNNscited by 441
  6. 2021Vector Neurons: A General Framework for SO(3)-Equivariant Networkscited by 304
  7. 2019Stochastic Latent Actor-Critic: Deep Reinforcement Learning with a Latent Variable Modelcited by 228
  8. 2019VariBAD: A Very Good Method for Bayes-Adaptive Deep RL via Meta-Learningcited by 210
  9. 2020MDP Homomorphic Networks: Group Symmetries in Reinforcement Learningcited by 133
  10. 2020Incorporating Symmetry into Deep Dynamics Models for Improved Generalizationcited by 123
  11. 2018Deep Variational Reinforcement Learning for POMDPscited by 122
  12. 2021Recurrent Model-Free RL Can Be a Strong Baseline for Many POMDPscited by 115
  13. 2016DESPOT: Online POMDP Planning with Regularizationcited by 105
  14. 2021Equivariant Point Network for 3D Point Cloud Analysiscited by 81
  15. 2022SO(2)-Equivariant Reinforcement Learningcited by 79
  16. 2015Memory-based control with recurrent neural networkscited by 77
  17. 2021Equivariant Q Learning in Spatial Action Spacescited by 61
  18. 2018Optimizing Agent Behavior over Long Time Scales by Transporting Valuecited by 57
  19. 2021Memory-based Deep Reinforcement Learning for POMDPscited by 53
  20. 2022Sample Efficient Grasp Learning Using Equivariant Modelscited by 52
  21. 2019Variational Recurrent Models for Solving Partially Observable Control Taskscited by 46
  22. 2022Equivariant Transporter Networkcited by 44
  23. 2023Neural Volumetric Memory for Visual Locomotion Controlcited by 40
  24. 2022On-Robot Learning With Equivariant Modelscited by 38
  25. 2022The Surprising Effectiveness of Equivariant Models in Domains with Latent Symmetrycited by 32
  26. 2020Discriminative Particle Filter Reinforcement Learning for Complex Partial Observationscited by 27
  27. 2021Unbiased Asymmetric Reinforcement Learning under Partial Observabilitycited by 24
  28. 2020Group Equivariant Deep Reinforcement Learningcited by 22
  29. 2021Recurrent Off-policy Baselines for Memory-based Continuous Controlcited by 19
  30. 2023A General Theory of Correct, Incorrect, and Extrinsic Equivariancecited by 18
  31. 2022Leveraging Fully Observable Policies for Learning under Partial Observabilitycited by 17
  32. 2022Equivariant Networks for Zero-Shot Coordinationcited by 16
  33. 2020Belief-Grounded Networks for Accelerated Robot Learning under Partial Observabilitycited by 9
  34. 2022BulletArm: An Open-Source Robotic Manipulation Benchmark and Learning Frameworkcited by 9
  35. 2022Hierarchical Reinforcement Learning under Mixed Observabilitycited by 4

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 1; 2025: 11; 2026: 5.

17 papers.

Sort:

By area:

  1. 2025Geometry-aware RL for Manipulation of Varying Shapes and Deformable Objectscited by 10
  2. 2025SE(3)-Equivariant Robot Learning and Control: A Tutorial Surveycited by 8
  3. 2024MATCH POLICY: A Simple Pipeline from Point Cloud Registration to Manipulation Policiescited by 6
  4. 2025Constraint-Preserving Data Generation for Visuomotor Policy Learningcited by 6
  5. 2025A Practical Guide for Incorporating Symmetry in Diffusion Policycited by 4
  6. 2025EfficientFlow: Efficient Equivariant Flow Policy Learning for Embodied AIcited by 4
  7. 2025Partially Equivariant Reinforcement Learning in Symmetry-Breaking Environmentscited by 3
  8. 2025Symmetries-enhanced Multi-Agent Reinforcement Learningcited by 1
  9. 2025Flow Equivariant Recurrent Neural Networkscited by 1
  10. 2026Conformal Orbit-Valid Trust Horizons for Equivariant World Modelscited by 1
  11. 2025Generalization Bounds for Equivariant Networks on Markov Datacited by 0
  12. 2025Multi-Group Equivariant Augmentation for Reinforcement Learning in Robot Manipulationcited by 0
5 more
  1. 2025Reinforcement Learning Using known Invariancescited by 0
  2. 2026Cost-Aware Diffusion Active Searchcited by 0
  3. 2026Reflex: Reinforcement Learning with Reflection Symmetry Exploitation in State-Based Continuous Controlcited by 0
  4. 2026Pix2Act: Image-Space Manipulation Policies with Equivariant Augmentationcited by 0
  5. 2026Terminal Symmetry as a Carrier of Asymmetric Process Knowledge: Statewise Refinement for Anytime Verified Constructioncited by 0