Research map / Multi-agent systems and games / Cooperative multi-agent RL
Multi-agent Reinforcement Learning for Networked System Control
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 |
|---|---|---|---|
| Use NeurComm instead of CommNet | A2C MARL · ATSC Grid | average episode return: -165.1 -> -136.1 | Helped |
| Use NeurComm instead of the strongest baseline | A2C MARL · ATSC Monaco | average episode return: -359.4 -> -226.3 | Helped |
| Use FPrint instead of NeurComm | A2C MARL · CACC Catch-up | average episode return: -94.6 -> -57.8 | Helped |
| Use FPrint instead of NeurComm | A2C MARL · CACC Slow-down | average episode return: -934.7 -> -697.9 | Helped |
| Tune the spatial discount factor for non-communicative MARL (ablation) | IA2C · ATSC and CACC scenarios | learning performance | Helped |
| Set spatial discount α to its scenario-specific best value (ablation) | IA2C · CACC Slow-down | learning performance | Helped |
| Set spatial discount α to its scenario-specific best value (ablation) | IA2C · CACC Catch-up | learning performance | No clear effect |
| Add encoded-input concatenation to the baseline communication protocol (ablation) | NeurComm · ATSC and CACC scenarios | learning performance | Helped |
| Add neighbor policy fingerprints to the baseline communication protocol (ablation) | NeurComm · ATSC and CACC scenarios | learning performance | Helped |
| Combine encoded-input concatenation and policy fingerprints (ablation) | NeurComm · ATSC and CACC scenarios | learning performance | Helped |
| Use NeurComm instead of existing communication protocols | A2C MARL · ATSC scenarios | sample efficiency and learning stability | Helped |
| Use delayed information sharing through NeurComm | A2C MARL · CACC scenarios | execution performance | Mixed |
Closest papers by meaning
- Multi-Agent Reinforcement Learning in Stochastic Networked Systems (NeurIPS 2021 · similarity 0.88)
- Fully Decentralized Multi-Agent Reinforcement Learning with Networked Agents (ICML 2018 · similarity 0.88)
- Value Propagation for Decentralized Networked Deep Multi-agent Reinforcement Learning (NeurIPS 2019 · similarity 0.88)
- Bayesian Ego-graph inference for Networked Multi-Agent Reinforcement Learning (NeurIPS 2025 · similarity 0.88)
- Communication-efficient Multi-Agent Reinforcement Learning with Spatiotemporal Information Hub (AAAI 2026 · similarity 0.87)
- Bridging Training and Execution via Dynamic Directed Graph-Based Communication in Cooperative Multi-Agent Systems (AAAI 2025 · 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.