Research map / Reinforcement learning / Policy gradient and actor-critic
Divergence-Augmented Policy Optimization
Qing Wang, Yingru Li, Jiechao Xiong, Tong Zhang
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 |
|---|---|---|---|
| add KL divergence augmentation to the return | PPO · Atari Qbert (58-game suite) | relative score improvement: +894% | Improved |
| add KL divergence augmentation to the return | PPO · Atari VideoPinball | relative score improvement: +781% | Improved |
| add KL divergence augmentation to the return | PPO · Atari Frostbite | relative score improvement: -19% | Hurt |
| add KL divergence augmentation to the return | PPO · Atari 58 games, hard exploration | relative score improvement | Improved |
| add KL divergence augmentation to the return | PPO · Atari Pong, Asterix, MontezumaRevenge | relative score improvement: 1%, 0% | No clear effect |
| replace divergence augmentation with entropy augmentation (ablation) | PPO+DA · Atari selected games | score | Hurt |
| tune entropy coefficient 1/eta over 0.5, 0.1, 0.01, 0.001 (ablation) | PPO+Entropy · Atari selected games | score: best at 0.1 | Mixed |
| use 1-step divergence instead of multi-step discounted divergence (c_D = 0) (ablation) | PPO+DA · Atari Enduro and Qbert | score | Hurt |
Closest papers by meaning
- Interpolated Policy Gradient: Merging On-Policy and Off-Policy Gradient Estimation for Deep Reinforcement Learning (NeurIPS 2017 · similarity 0.89)
- Trajectory-Based Off-Policy Deep Reinforcement Learning (ICML 2019 · similarity 0.89)
- Adaptive Advantage-Guided Policy Regularization for Offline Reinforcement Learning (ICML 2024 · similarity 0.88)
- Relaxed Stationary Distribution Correction Estimation for Improved Offline Policy Optimization (AAAI 2024 · similarity 0.88)
- Efficient Off-Policy Learning for High-Dimensional Action Spaces (ICLR 2025 · similarity 0.87)
- Seizing Serendipity: Exploiting the Value of Past Success in Off-Policy Actor-Critic (ICML 2024 · 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.
