Research map / Multi-agent systems and games / Cooperative multi-agent RL
Stabilising Experience Replay for Deep Multi-Agent Reinforcement Learning
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 |
|---|---|---|---|
| Add training episode and exploration rate fingerprints to experience replay | Feed-forward IQL · StarCraft micromanagement | performance | Helped |
| Apply multi-agent importance sampling to experience replay | Feed-forward IQL · StarCraft micromanagement | performance | Helped |
| Combine importance sampling with fingerprints | Experience replay with fingerprints · StarCraft micromanagement | performance | No clear effect |
| Add fingerprints to recurrent experience replay | Recurrent IQL · StarCraft m3v3 | performance | No clear effect |
| Apply importance sampling to recurrent experience replay | Recurrent IQL · StarCraft m3v3 | performance | No clear effect |
| Use recurrent experience replay rather than no replay | Recurrent IQL · StarCraft micromanagement | performance | Helped |
| Use experience replay rather than no replay | Feed-forward IQL · StarCraft micromanagement | performance | Helped |
| Apply importance sampling to experience replay | Feed-forward IQL · StarCraft m5v5 | performance | Helped |
| Add fingerprints to experience replay | IQL · StarCraft m5v5 | performance | Helped |
| Add fingerprints to experience replay | Recurrent IQL · StarCraft micromanagement | performance | Mixed |
Closest papers by meaning
- Actor Prioritized Experience Replay (Abstract Reprint) (AAAI 2024 · similarity 0.89)
- Large Batch Experience Replay (ICML 2022 · similarity 0.88)
- Selectively Sharing Experiences Improves Multi-Agent Reinforcement Learning (NeurIPS 2023 · similarity 0.88)
- Revisiting Fundamentals of Experience Replay (ICML 2020 · similarity 0.87)
- Novelty-Guided Data Reuse for Efficient and Diversified Multi-Agent Reinforcement Learning (AAAI 2025 · similarity 0.87)
- A Generalist Hanabi Agent (ICLR 2025 · similarity 0.85)
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.