Research map / Multi-agent systems and games / Cooperative multi-agent RL
Automatic Grouping for Efficient Cooperative Multi-Agent Reinforcement Learning
Yifan Zang, Jinmin He, Kai Li, Haobo Fu, QIANG FU, Junliang Xing, Jian Cheng
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 |
|---|---|---|---|
| replace learned dynamic grouping with all agents in one group (ablation) | GoMARL · SMAC corridor | test win rate | Hurt |
| replace learned grouping with each agent as its own group or equal split (ablation) | GoMARL · SMAC corridor | test win rate | Hurt |
| replace dynamic grouping with natural type-based grouping (ablation) | GoMARL · SMAC MMM2 | test win rate | Hurt |
| replace learned grouping with homogeneous-type grouping (ablation) | GoMARL · SMAC 3s5z_vs_3s6z | test win rate | No clear effect |
| add specialized agent networks (SAN) to baselines (ablation) | QMIX/QPLEX/VAST/RIIT · SMAC maps | test win rate | Improved |
| replace group state sg with full state s (ablation) | GoMARL · SMAC super hard maps | test win rate | Hurt |
| remove group-related info e (vanilla agent networks) (ablation) | GoMARL · SMAC super hard maps | test win rate | Hurt |
| dual-hierarchical mixing vs single monotonic mixing network | GoMARL · SMAC 8m_vs_9m | mixing network parameters: 53.313K -> 51.427K | Improved |
| dual-hierarchical mixing vs single monotonic mixing network | GoMARL · SMAC 3s_vs_5z | mixing network parameters: 21.601K -> 26.530K | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- QPLEX — QPLEX: Duplex Dueling Multi-Agent Q-Learning (ICLR 2021) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Group-in-Group Policy Optimization for LLM Agent Training (NeurIPS 2025 · similarity 0.89)
- N-agent Ad Hoc Teamwork (NeurIPS 2024 · similarity 0.89)
- Multi-agent cooperation through learning-aware policy gradients (ICLR 2025 · similarity 0.88)
- Group-Graph Policy Optimization for Long-Horizon Agentic Reinforcement Learning (NeurIPS 2026 · similarity 0.88)
- Stronger-MAS: Multi-Agent Reinforcement Learning for Collaborative LLMs (ICLR 2026 · similarity 0.88)
- LLM-Assisted Semantically Diverse Teammate Generation for Efficient Multi-agent Coordination (ICML 2025 · similarity 0.88)
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.
