Research map / Multi-agent systems and games / Cooperative multi-agent RL
A Unified Framework for Factorizing Distributional Value Functions for Multi-Agent Reinforcement Learning
Wei-Fang Sun, Cheng-Kuang Lee, Simon See, Chun-Yi Lee
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 expected-value learner with distributional IQN learner (DIQL) | IQL · SMAC 3s5z_vs_3s6z | median win rate (%): 29.83 -> 62.22 | Improved |
| replace expected-value learner with distributional IQN learner (DIQL) | IQL · SMAC 27m_vs_30m | median win rate (%): 2.27 -> 6.02 | Improved |
| replace expected-value learner with distributional IQN learner (DIQL) | IQL · SMAC corridor | median win rate (%): 84.87 -> 91.62 | Improved |
| replace expected-value learner with distributional IQN learner (DIQL) | IQL · SMAC 6h_vs_8z | median win rate (%): 0.00 -> 0.00 | No clear effect |
| add DFAC (distributional factorization, DMIX) | QMIX · SMAC 6h_vs_8z | median win rate (%): 12.78 -> 49.43 | Improved |
| add DFAC (distributional factorization, DMIX) | QMIX · SMAC corridor | median win rate (%): 37.61 -> 90.45 | Improved |
| add DFAC (distributional factorization, DPLEX) | QPLEX · SMAC 6h_vs_8z | median win rate (%): 0.00 -> 43.75 | Improved |
| add DFAC (distributional factorization, DPLEX) | QPLEX · SMAC MMM2 | median win rate (%): 96.88 -> 96.88 | No clear effect |
| add DFAC (distributional factorization, DPLEX) | QPLEX · SMAC 27m_vs_30m | median win rate (%): 78.12 -> 90.62 | Improved |
| add DFAC (DDN, distributional VDN with IQN) | VDN · SMAC 6h_vs_8z | median win rate (%): 0.00 -> 83.92 | Improved |
| add DFAC (distributional factorization, DMIX) | QMIX · SMAC Ultra Hard 26m_vs_30m | median win rate (%): 62.78 -> 81.82 | Improved |
| add DFAC (distributional factorization, DPLEX) | QPLEX · SMAC Ultra Hard 26m_vs_30m | median win rate (%): 78.12 -> 59.38 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- QPLEX — QPLEX: Duplex Dueling Multi-Agent Q-Learning (ICLR 2021) · better on 6 of 6 shared results
- QMIX — QMIX: Monotonic Value Function Factorisation for Deep Multi-Agent Reinforcement Learning (ICML 2018) · better on 5 of 5 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
- DFAC Framework: Factorizing the Value Function via Quantile Mixture for Multi-Agent Distributional Q-Learning (ICML 2021 · similarity 0.96)
- ResQ: A Residual Q Function-based Approach for Multi-Agent Reinforcement Learning Value Factorization (NeurIPS 2022 · similarity 0.91)
- Towards Understanding Cooperative Multi-Agent Q-Learning with Value Factorization (NeurIPS 2021 · similarity 0.91)
- PAC: Assisted Value Factorization with Counterfactual Predictions in Multi-Agent Reinforcement Learning (NeurIPS 2022 · similarity 0.91)
- Factored Value Functions for Graph-Based Multi-Agent Reinforcement Learning (ICML 2026 · similarity 0.90)
- ConcaveQ: Non-monotonic Value Function Factorization via Concave Representations in Deep Multi-Agent Reinforcement Learning (AAAI 2024 · 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.
