mlsanityData to 14 Oct 2026

Reflex: Reinforcement Learning with Reflection Symmetry Exploitation in State-Based Continuous Control

arXiv 2605.23415 ↗First posted 22 May 2026cs.LG cs.AIcited by 0

On the map: Deep reinforcement learning methods → RL for robot locomotion and manipulation

Abstract (arXiv)

Reinforcement learning has long struggled with poor sample efficiency. One promising approach to mitigate this problem is leveraging group-invariant Markov Decision Processes (G-invariant MDPs). Existing works in this direction have primarily focused on image-based RL and rotational symmetry such as SO(2), leaving state-based RL and reflection symmetry largely underexplored. In this work, we focus on state-based continuous control tasks and exploit reflection symmetry by introducing Reflex, a paradigm that seamlessly integrates with both on-policy and off-policy RL algorithms. We formalize two types of reflection-axial reflection and bilateral reflection, and characterize their corresponding transformations. Building on a theoretical analysis of symmetry-preserving optimal value functions and policies, Reflex integrates reflection symmetry into policy learning through principled symmetry regularization mechanisms. We integrate Reflex with PPO and SAC, and evaluate it on a suite of OpenAI Gym and DeepMind Control benchmarks, demonstrating superior performance over standard baselines while improving sample efficiency. Our code is available at https://github.com/TonyStark042/Reflex.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in RL for robot locomotion and manipulation, one of 15 directions in Deep reinforcement learning methods. Builds on 31 of its 41 references that are in the map. First posted on arXiv 22 May 2026.

Builds on

31 of 41 references.

Closest ancestors · 4

Same direction, newest first

  1. 2024Reinforcement Learning with Euclidean Data Augmentation for State-Based Continuous Control16 Oct
  2. 2024Equivariant Reinforcement Learning under Partial Observability26 Aug
  3. 2024Leveraging Symmetry in RL-based Legged Locomotion Control26 Mar
  4. 2024Revisiting Data Augmentation in Deep Reinforcement Learning19 Feb

Other references · 27

  1. 2017Proximal Policy Optimization Algorithmscited by 15,215
  2. 2015Continuous control with deep reinforcement learningcited by 3,407
  3. 2018Addressing Function Approximation Error in Actor-Critic Methodscited by 2,417
  4. 2015High-Dimensional Continuous Control Using Generalized Advantage Estimationcited by 2,026
  5. 2016OpenAI Gymcited by 1,736
  6. 2018Soft Actor-Critic Algorithms and Applicationscited by 1,240
  7. 2016Group Equivariant Convolutional Networkscited by 1,165
  8. 2020CURL: Contrastive Unsupervised Representations for Reinforcement Learningcited by 760
  9. 2018QT-Opt: Scalable Deep Reinforcement Learning for Vision-Based Robotic Manipulationcited by 743
  10. 2020Image Augmentation Is All You Need: Regularizing Deep Reinforcement Learning from Pixelscited by 585
  11. 2020Reinforcement Learning with Augmented Datacited by 468
  12. 2020dm_control: Software and Tasks for Continuous Controlcited by 375
15 more
  1. 2021Mastering Visual Continuous Control: Improved Data-Augmented Reinforcement Learningcited by 372
  2. 2020Transporter Networks: Rearranging the Visual World for Robotic Manipulationcited by 344
  3. 2021A Practical Method for Constructing Equivariant Multilayer Perceptrons for Arbitrary Matrix Groupscited by 187
  4. 2020MDP Homomorphic Networks: Group Symmetries in Reinforcement Learningcited by 142
  5. 2020Counterfactual Data Augmentation using Locally Factored Dynamicscited by 93
  6. 2022SO(2)-Equivariant Reinforcement Learningcited by 84
  7. 2021Geometric Deep Learning and Equivariant Neural Networkscited by 65
  8. 2018Learning Symmetric and Low-energy Locomotioncited by 61
  9. 2022MoCoDA: Model-based Counterfactual Data Augmentationcited by 44
  10. 2024Symmetry Considerations for Learning Task Symmetric Robot Policiescited by 28
  11. 2024Morphological Symmetries in Roboticscited by 13
  12. 2023Guided Data Augmentation for Offline Reinforcement Learning and Imitation Learningcited by 10
  13. 2023Understanding when Dynamics-Invariant Data Augmentations Benefit Model-Free Reinforcement Learning Updatescited by 7
  14. 2023Addressing imperfect symmetry: A novel symmetry-learning actor-critic extensioncited by 5
  15. 2024Equivariant Offline Reinforcement Learningcited by 5

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Sample Complexity of Equivariant Reinforcement Learning
  2. 2025Partially Equivariant Reinforcement Learning in Symmetry-Breaking Environments
  3. 2024Equivariant Action Sampling for Reinforcement Learning and Planning
  4. 2024Exploiting Symmetry in Dynamics for Model-Based Reinforcement Learning with Asymmetric Rewards
  5. 2023Look Beneath the Surface: Exploiting Fundamental Symmetry for Sample-Efficient Offline RL
  6. 2026MIRROR: Reflecting Hidden Symmetries from Offline Robot Interaction Data ↗
  7. 2024Leveraging Symmetry to Accelerate Learning of Trajectory Tracking Controllers for Free-Flying Robotic Systems
  8. 2026Beyond State-Wise Mirror Descent: Offline Policy Optimization with Parametric Policies