Research map / Reinforcement learning / Policy gradient and actor-critic
Policy Gradient Methods in the Presence of Symmetries and State Abstractions
Prakash Panangaden, Sahand Rezaei-Shoshtari, Rosie Zhao, David Meger, Doina Precup
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| sum HPG and DPG gradients instead of independent updates (ablation) | Deterministic DHPG · DeepMind Control Suite (pixels) | gradient variance | Improved |
| drop the DPG update and use only HPG (ablation) | Deterministic DHPG · DeepMind Control Suite, cartpole and pendulum | return | No clear effect |
| add the DPG update to HPG (ablation) | Deterministic DHPG · DeepMind Control Suite, walker | return | Improved |
| use a single critic for actual and abstract MDPs (ablation) | Deterministic DHPG · DeepMind Control Suite | sample efficiency | Mixed |
| use 3-step instead of 1-step return (ablation) | Deterministic DHPG · DeepMind Control Suite with augmentation | early training speed | Improved |
| use 3-step instead of 1-step return (ablation) | Deterministic DHPG · DeepMind Control Suite without augmentation | final performance | Hurt |
Closest papers by meaning
- Continuous MDP Homomorphisms and Homomorphic Policy Gradient (NeurIPS 2022 · similarity 0.96)
- Learning Markov State Abstractions for Deep Reinforcement Learning (NeurIPS 2021 · similarity 0.90)
- Delayed homomorphic reinforcement learning for environments with delayed feedback (NeurIPS 2026 · similarity 0.89)
- ${\rm E}(3)$-Equivariant Actor-Critic Methods for Cooperative Multi-Agent Reinforcement Learning (ICML 2024 · similarity 0.88)
- Wasserstein Policy Optimization (ICML 2025 · similarity 0.88)
- Deterministic Policy Gradient Primal-Dual Methods for Continuous-Space Constrained MDPs (AAAI 2025 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
