Research map / Reinforcement learning / Policy gradient and actor-critic
Reinforcement Learning with Deep Energy-Based Policies
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| use a maximum-entropy energy-based policy with amortized SVGD sampling instead of deterministic DDPG | Soft Q-learning · Swimmer snake | exploration of forward and backward modes before discovering the finish-line rew: DDPG: 80% of policies converged to forward motion; 20% chose the suboptimal backward mode | Helped |
| use a maximum-entropy energy-based policy with amortized SVGD sampling instead of deterministic DDPG | Soft Q-learning · Quadrupedal robot maze | exploration success: DDPG: only 60% of policies converged to choosing the lower passage | Helped |
| pretrain with a maximum-entropy policy that moves in arbitrary directions instead of deterministic DDPG pretraining | Quadrupedal robot policy fine-tuning · Hallway and U-shaped maze tasks | average discounted return during fine-tuning | Helped |
| reduce the entropy coefficient α during pretraining (ablation) | Soft Q-learning · Quadrupedal robot pretraining | behavior diversity | Mixed |
Closest papers by meaning
- S$2$AC: Energy-Based Reinforcement Learning with Stein Soft Actor Critic (ICLR 2024 · similarity 0.93)
- Maximum Entropy Reinforcement Learning via Energy-Based Normalizing Flow (NeurIPS 2024 · similarity 0.89)
- DIME: Diffusion-Based Maximum Entropy Reinforcement Learning (ICML 2025 · similarity 0.87)
- Generative Actor-Critic with Soft Bridge Policies (NeurIPS 2026 · similarity 0.87)
- Maximum Entropy Reinforcement Learning with Diffusion Policy (ICML 2025 · similarity 0.86)
- Stabilizing Reinforcement Learning in Differentiable Multiphysics Simulation (ICLR 2025 · similarity 0.86)
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.