Research map / Multi-agent systems and games / Cooperative multi-agent RL
Learning Fairness in Multi-Agent Systems
Jiechuan Jiang, Zongqing Lu
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 FEN's fair-efficient reward and hierarchical controller with specialized sub-policies instead of Independent agents | Multi-agent system · Job scheduling | CV: 1.57 -> 0.17 | Improved |
| Use FEN instead of Independent agents | Multi-agent system · Job scheduling | resource utilization: 96% -> 90% | Hurt |
| Use FEN instead of Min+αAvg | Multi-agent system · Job scheduling | CV: 0.39 -> 0.17 | Improved |
| Remove the hierarchy (ablation) | FEN · Job scheduling | resource utilization: 90% -> 57% | Hurt |
| Add the hierarchy (ablation) | FEN · Job scheduling | CV: 0.22 -> 0.17 | Improved |
| Add the hierarchy (ablation) | FEN · The Matthew effect | episodes to converge: 6000 -> 300 | Improved |
| Use FEN instead of Independent agents | Multi-agent system · The Matthew effect | social welfare: 791 -> 830 | Improved |
| Use FEN instead of Independent agents | Multi-agent system · The Matthew effect | CV: 0.86 -> 0.06 | Improved |
| Estimate average utility with local gossip consensus (ablation) | FEN · The Matthew effect | CV: 0.06 -> 0.07 | No clear effect |
| Replace learned fairness sub-policies with a random sub-policy (ablation) | FEN · The Matthew effect | CV: 0.06 -> 0.08 | Hurt |
| Use FEN instead of Min+αAvg | Multi-agent system · Manufacturing plant | no. products: 23 -> 48 | Improved |
| Use FEN instead of Min+αAvg | Multi-agent system · Manufacturing plant | resource utilization: 34% -> 82% | Improved |
Closest papers by meaning
- Learning Fair Policies in Decentralized Cooperative Multi-Agent Reinforcement Learning (ICML 2021 · similarity 0.90)
- Fair Cooperation in Mixed-Motive Games via Conflict-Aware Gradient Adjustment (NeurIPS 2025 · similarity 0.88)
- FairDICE: Fairness-Driven Offline Multi-Objective Reinforcement Learning (NeurIPS 2025 · similarity 0.87)
- Controlling Behavioral Diversity in Multi-Agent Reinforcement Learning (ICML 2024 · similarity 0.85)
- Intersectional Fairness in Reinforcement Learning with Large State and Constraint Spaces (ICML 2025 · similarity 0.85)
- HyperMARL: Adaptive Hypernetworks for Multi-Agent RL (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.
