Research map / Reinforcement learning / Policy gradient and actor-critic
Tracking Drift: Variation-Aware Entropy Scheduling for Non-Stationary Reinforcement Learning
Tongxi Wang, Zhuoyang Xia, Xinran Chen, Shan Liu
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 AES entropy scheduling | SAC · MuJoCo (abrupt) | nAUC: 0.67 -> 0.87 | Improved |
| add AES entropy scheduling | PPO · MuJoCo (abrupt) | nAUC: 0.66 -> 0.69 | No clear effect |
| add AES entropy scheduling | MEow · MuJoCo (abrupt) | nAUC: 0.67 -> 0.95 | Improved |
| add AES entropy scheduling | SQL · Isaac Gym (abrupt) | nAUC: 0.39 -> 0.66 | Improved |
| add AES entropy scheduling | SAC · Toy (steady) | nAUC: 1.00 -> 1.13 | Improved |
| add AES entropy scheduling | SQL · MuJoCo (steady) | nAUC: 0.80 -> 0.81 | No clear effect |
| add AES entropy scheduling | PPO · Toy (steady) | nAUC: 0.89 -> 0.89 | No clear effect |
| add AES entropy scheduling | SAC · Abrupt-change recovery time | recovery time (%): 13.96 -> 7.74 | Improved |
| add AES entropy scheduling | PPO · Abrupt-change recovery time | recovery time (%): 12.12 -> 8.43 | Improved |
| add AES entropy scheduling | SQL · Abrupt-change recovery time | recovery time (%): 15.73 -> 9.73 | Improved |
| add AES entropy scheduling | MEow · Abrupt-change recovery time | recovery time (%): 11.58 -> 6.42 | Improved |
| add AES entropy scheduling | MEow · MuJoCo (mixed) | drop area ratio: 0.32 -> 0.09 | Improved |
Closest papers by meaning
- AEM: Adaptive Entropy Modulation for Multi-Turn Agentic Reinforcement Learning (NeurIPS 2026 · similarity 0.89)
- State Entropy Regularization for Robust Reinforcement Learning (NeurIPS 2025 · similarity 0.87)
- Scalable Maximum Entropy Reinforcement Learning for Diffusion Policies via Adjoint Matching (NeurIPS 2026 · similarity 0.87)
- Pausing Policy Learning in Non-stationary Reinforcement Learning (ICML 2024 · similarity 0.87)
- DIME: Diffusion-Based Maximum Entropy Reinforcement Learning (ICML 2025 · similarity 0.87)
- Behavioral Entropy-Guided Dataset Generation for Offline Reinforcement Learning (ICLR 2025 · similarity 0.87)
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.
