Research map / Multi-agent systems and games / Cooperative multi-agent RL
Distributed Influence-Augmented Local Simulators for Parallel MARL in Large Networked Systems
Miguel Suau, Jinke He, Mustafa Mert Celikok, Matthijs T. J. Spaan, Frans A Oliehoek
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 |
|---|---|---|---|
| train with DIALS instead of global simulator | MARL · 100-agent traffic | training time: more than 10 days -> less than 6 hours | Improved |
| train with DIALS instead of global simulator | MARL · 4-agent traffic and warehouse | average return | Improved |
| use untrained influence predictors (ablation) | DIALS · traffic and warehouse | average return | Hurt |
| retrain AIPs only at start (F=4M) instead of every 1M (ablation) | DIALS · traffic 25 agents | average return | Hurt |
| retrain AIPs only once (F=4M) (ablation) | DIALS · warehouse 25 agents | average return | No clear effect |
| retrain AIPs too frequently (F=100K) (ablation) | DIALS · warehouse 25 agents | average return | Hurt |
| increase number of simulators/processes | DIALS · memory usage | total memory | Hurt |
Closest papers by meaning
- Influence-Augmented Local Simulators: a Scalable Solution for Fast Deep RL in Large Networked Systems (ICML 2022 · similarity 0.92)
- Multi-Agent Reinforcement Learning in Stochastic Networked Systems (NeurIPS 2021 · similarity 0.87)
- Bayesian Ego-graph inference for Networked Multi-Agent Reinforcement Learning (NeurIPS 2025 · similarity 0.87)
- Bridging Training and Execution via Dynamic Directed Graph-Based Communication in Cooperative Multi-Agent Systems (AAAI 2025 · similarity 0.86)
- ComaDICE: Offline Cooperative Multi-Agent Reinforcement Learning with Stationary Distribution Shift Regularization (ICLR 2025 · similarity 0.85)
- Towards Efficient Collaboration via Graph Modeling in Reinforcement Learning (AAAI 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.
