PRISM: Parallel Reward Integration with Symmetry for MORL
Not on the map. Why?
Abstract (arXiv)
This work studies heterogeneous Multi-Objective Reinforcement Learning (MORL), where objectives can differ sharply in temporal frequency. Such heterogeneity allows dense objectives to dominate learning, while sparse long-horizon rewards receive weak credit assignment, leading to poor sample efficiency. We propose a Parallel Reward Integration with Symmetry (PRISM) algorithm that enforces reflectional symmetry as an inductive bias in aligning reward channels. PRISM introduces ReSymNet, a theory-motivated model that reconciles temporal-frequency mismatches across objectives, using residual blocks to learn a scaled opportunity value that accelerates exploration while preserving the optimal policy. We also propose SymReg, a reflectional equivariance regulariser that enforces agent mirroring and constrains policy search to a reflection-equivariant subspace. This restriction provably reduces hypothesis complexity and improves generalisation. Across MuJoCo benchmarks, PRISM consistently outperforms both a sparse-reward baseline and an oracle trained with full dense rewards, improving Pareto coverage and distributional balance: it achieves hypervolume gains exceeding 100% over the baseline and up to 32% over the oracle. The code is at hrefhttps://github.com/EVIEHub/PRISMhttps://github.com/EVIEHub/PRISM.
Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 28 of its 50 references that are in the map. First posted on arXiv 20 Feb 2026.
Builds on
28 of 50 references.
References · 28
- 2015Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classificationcited by 3,590
- 2017Curiosity-driven Exploration by Self-supervised Predictioncited by 1,173
- 2018Exploration by Random Network Distillationcited by 966
- 2020Deep Reinforcement Learning for Autonomous Driving: A Surveycited by 698
- 2017Spectrally-normalized margin bounds for neural networkscited by 681
- 2018A Survey of Inverse Reinforcement Learning: Challenges, Methods and Progresscited by 294
- 2021A Practical Guide to Multi-Objective Reinforcement Learning and Planningcited by 240
- 2019A Generalized Algorithm for Multi-Objective Reinforcement Learning and Policy Adaptationcited by 166
- 2024Deep Reinforcement Learning for Robotics: A Survey of Real-World Successescited by 153
- 2020MDP Homomorphic Networks: Group Symmetries in Reinforcement Learningcited by 133
- 2021Recent Advances in Reinforcement Learning in Financecited by 102
- 2019A survey on intrinsic motivation in reinforcement learningcited by 79
16 more
- 2022SO(2)-Equivariant Reinforcement Learningcited by 79
- 2021Equivariant Q Learning in Spatial Action Spacescited by 61
- 2023Sample-Efficient Multi-Objective Learning via Generalized Policy Improvement Prioritizationcited by 46
- 2019Why ResNet Works? Residuals Generalizecited by 39
- 2021Self-Supervised Online Reward Shaping in Sparse-Reward Environmentscited by 34
- 2018Meta-Learning for Multi-objective Reinforcement Learningcited by 33
- 2021Learning Long-Term Reward Redistribution via Randomized Return Decompositioncited by 33
- 2020Learning Guidance Rewards with Trajectory-space Smoothingcited by 30
- 2019On the Generalization Gap in Reparameterizable Reinforcement Learningcited by 21
- 2025Pareto Set Learning for Multi-Objective Reinforcement Learningcited by 18
- 2024Approximate Equivariance in Reinforcement Learningcited by 9
- 2022Benefits of Permutation-Equivariance in Auction Mechanismscited by 7
- 2025Preference-based Multi-Objective Reinforcement Learningcited by 7
- 2025Attention-Based Reward Shaping for Sparse and Delayed Rewardscited by 5
- 2024Beyond Simple Sum of Delayed Rewards: Non-Markovian Reward Modeling for Reinforcement Learningcited by 2
- 2025Multi-Objective Reinforcement Learning for Power Grid Topology Controlcited by 1
Built on by
Not cited yet.