Research map / Multi-agent systems and games / Games and equilibria
Decentralised Learning in Systems With Many, Many Strategic Agents
David Mguni, Joel Jennings, Enrique Munoz de Cote
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 |
|---|---|---|---|
| Use actor-critic fictitious play with a mean-field formulation | Large-population spatial congestion game · Spatial Congestion Game | policy and reward stability: after 2,000 episodes | Improved |
| Increase the congestion averseness parameter alpha | Mean-field actor-critic fictitious play · Spatial Congestion Game | agent dispersion: alpha in {1.0, 2.0, 3.0} | Improved |
| Use the mean-field actor-critic method for a bimodal reward landscape | Spatial Congestion Game · Bimodal Gaussian rewards | distribution across reward peaks | Improved |
| Include non-zero movement costs and learn a long-term route | Mean-field actor-critic fictitious play · Supply with Uncertain Demand | cumulative payoff: eta=2 | Improved |
| Set movement cost eta to zero | Mean-field actor-critic fictitious play · Supply with Uncertain Demand | route tracking: eta=0 | Mixed |
| Use actor-critic fictitious play to learn the mean-field LQC policy | Mean-Field Linear Quadratic Regulator · Mean-Field Linear Quadratic Regulator | mu: (0.50000, -0.50000) -> (0.50717, -0.50537) | Improved |
| Use actor-critic fictitious play to learn the mean-field LQC policy | Mean-Field Linear Quadratic Regulator · Mean-Field Linear Quadratic Regulator | sigma^2: 0.14060 -> 0.16100 | Improved |
Closest papers by meaning
- Decentralized Q-learning in Zero-sum Markov Games (NeurIPS 2021 · similarity 0.91)
- Decentralized Reinforcement Learning: Global Decision-Making via Local Economic Transactions (ICML 2020 · similarity 0.90)
- Stability of Multi-Agent Learning in Competitive Networks: Delaying the Onset of Chaos (AAAI 2024 · similarity 0.89)
- Multi-Agent Learning under Uncertainty: Recurrence vs. Concentration (NeurIPS 2025 · similarity 0.89)
- Independent Learning of Nash Equilibria in Partially Observable Markov Potential Games with Decoupled Dynamics (NeurIPS 2026 · similarity 0.88)
- Asymptotic Extinction in Large Coordination Games (AAAI 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.
