Research map / Multi-agent systems and games / Cooperative multi-agent RL
Coordinated Proximal Policy Optimization
Zifan Wu, Chao Yu, Deheng Ye, Junge Zhang, haiyin piao, Hankz Hankui Zhuo
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 |
|---|---|---|---|
| use CoPPO coordinated step-size adaptation instead of independent PPO updates (MAPPO) | CoPPO · 4-agent 9-action penalty matrix game | average reward | Improved |
| remove the double clipping trick (ablation) | CoPPO · penalty matrix game | average reward and gradient variance | Hurt |
| vary inner clipping threshold among 0.05, 0.10, 0.15 (ablation) | CoPPO · penalty matrix game | average reward | Mixed |
| remove double clipping (ablation) | CoPPO · SMAC (two maps) | win rate | Hurt |
| replace weighted sum of local advantages with non-negative mixing network (Amix vs Asum) (ablation) | CoPPO · SMAC (two maps) | win rate | No clear effect |
| use CoPPO instead of MAPPO, COMA, DOP, QMIX, QTRAN | CoPPO · SMAC (3s5z, 10m_vs_11m, MMM2) | win rate | Improved |
| CoPPO instead of baselines across six matrix games | CoPPO · six 4-agent matrix games | average reward | Improved |
Closest papers by meaning
- The Surprising Effectiveness of PPO in Cooperative Multi-Agent Games (NeurIPS 2022 · similarity 0.91)
- Local Optimization Achieves Global Optimality in Multi-Agent Reinforcement Learning (ICML 2023 · similarity 0.89)
- Learning Explicit Credit Assignment for Cooperative Multi-Agent Reinforcement Learning via Polarization Policy Gradient (AAAI 2023 · similarity 0.88)
- Simple Policy Optimization (ICML 2025 · similarity 0.86)
- Scalable Constrained Policy Optimization for Safe Multi-agent Reinforcement Learning (NeurIPS 2024 · similarity 0.86)
- Effective Policy Learning for Multi-Agent Online Coordination Beyond Submodular Objectives (NeurIPS 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.
