Research map / Multi-agent systems and games / Cooperative multi-agent RL
Value factorization
Decomposes joint Q-functions into per-agent values with theory and improved factorization schemes.
39 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2018–2026) · an approach in Cooperative multi-agent RL, keyword cluster “decentralized · value · joint”
State History Tried Checklist Preprints Who Export
Where it stands
Not enough benchmark data Fewer than four papers report the same benchmark, so there is no record to follow.
10 papers in 2023–24 and 6 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic shrank, from 7.1% to 2.6%.
How this is decided
A benchmark is tracked when at least four papers of this approach report their own method on the same dataset and metric in at least two different years. Its record is the best number reported up to 2024 compared with the best reported in 2025–26: still improving means at least 1% better, flat means at most 0.2% better with at least two papers from 2025–26 trying. Numbers on a different scale from the rest (a fraction among percentages) and far outliers are left out. Results come from different papers with different setups, so they are signals, not head-to-head comparisons. The share of papers is counted at ICML, NeurIPS, ICLR, CVPR and AAAI.
How it got here
Most cited paper of each year
- 2018 · QMIX: Monotonic Value Function Factorisation for Deep Multi-Agent Reinforcement Learning (ICML 2018 · 343 citations)
- 2019 · QTRAN: Learning to Factorize with Transformation for Cooperative Multi-Agent Reinforcement Learning (ICML 2019 · 213 citations)
- 2020 · Shapley Q-Value: A Local Reward Approach to Solve Global Reward Games (AAAI 2020 · 74 citations)
- 2021 · QPLEX: Duplex Dueling Multi-Agent Q-Learning (ICLR 2021 · 150 citations)
- 2022 · SHAQ: Incorporating Shapley Value Theory into Multi-Agent Q-Learning (NeurIPS 2022 · 9 citations)
- 2023 · Contrastive Identity-Aware Learning for Multi-Agent Value Decomposition (AAAI 2023 · 26 citations)
- 2024 · ConcaveQ: Non-monotonic Value Function Factorization via Concave Representations in Deep Multi-Agent Reinforcement Learning (AAAI 2024 · 8 citations)
- 2025 · Multi-agent Markov Entanglement (NeurIPS 2025 · 0 citations)
- 2026 · Beyond Monotonicity: Revisiting Factorization Principles in Multi-Agent Q-Learning (AAAI 2026 · 1 citation)
Earliest papers here
- QMIX: Monotonic Value Function Factorisation for Deep Multi-Agent Reinforcement Learning (ICML 2018 · 343 citations)
- QTRAN: Learning to Factorize with Transformation for Cooperative Multi-Agent Reinforcement Learning (ICML 2019 · 213 citations)
- MAVEN: Multi-Agent Variational Exploration (NeurIPS 2019 · 67 citations)
What has been tried
32 papers here report 280 measured changes (a component added, removed or swapped, compared with a baseline): helped 58%, no clear effect 13%, hurt 23%, mixed 6%. Papers mostly publish changes that helped, so "helped" is over-represented.
Ablations in recent papers
| What was changed | Measured on | Result | Effect | Paper |
|---|---|---|---|---|
| Replace the concave mixing network with a 4-layer monotonic network | SMAC 3s_vs_5z · reward | Hurt | ConcaveQ: Non-monotonic Value Function Factorization via Concave Representations in Deep Multi-Agent Reinforcement Learning (AAAI 2024 · 8 citations) | |
| Add the temporal-aware preference attention layer | OpenIn., CloseOut., 3s5z, 6h vs 8z · accumulated trajectory reward | 23.43±0.95 -> 292.16±11.38; 718.16±22.38 -> 786.70±26.03; 8.32±0.38 -> 19.59±0.42; 13.26±0.31 -> 17.05±0.53 | Helped | Decoding Global Preferences: Temporal and Cooperative Dependency Modeling in Multi-Agent Preference-Based Reinforcement Learning (AAAI 2024 · 8 citations) |
| Use a small complement factor β rather than a value close to 1 | Evaluated tasks · performance | Helped | Optimistic Value Instructors for Cooperative Multi-Agent Reinforcement Learning (AAAI 2024 · 2 citations) | |
| Apply ReBorn only to the mixing network, preserving the Knowledge Invariant principle | SMAC: MMM2 · test win rate | Helped | The Dormant Neuron Phenomenon in Multi-Agent Reinforcement Learning Value Factorization (NeurIPS 2024 · 1 citation) |
What reviewers will expect
Datasets and baselines named in the 8 papers of this approach from 2024–2026 that have a full-text summary. Leaving out the common ones is a frequent review question.
Datasets
- SMAC · 3 of 8 papers
- Predator-Prey · 2 of 8 papers
Baselines
- QMIX · 3 of 8 papers
- QPLEX · 3 of 8 papers
- MAGAIL · 2 of 8 papers
Recent preprints
Loading recent arXiv preprints…
Who publishes on it
Institutions on the most papers of this approach since 2023 (of 16 papers; from OpenAlex and the papers' affiliations).
- Zhejiang University · 2 papers
- Nanjing University · 2 papers
- Alibaba Group · 2 papers
Use it
Download BibTeX of 30 key papersOpen the topic in the interactive map
Get alerts. Describe your idea in scime atlas and save it (free scime account): you get a weekly email when new preprints land close to it.
Related approaches
Same keyword cluster
- Decentralized actor-critic MARL (77 papers)
- Offline and networked MARL (65 papers)
- Coordination graphs (50 papers)
Elsewhere in the topic
- Cooperative MARL algorithms (149 papers)
- MARL benchmarks and grouping (121 papers)
- Learned inter-agent communication (58 papers)
- Intrinsic-reward exploration for MARL (34 papers)
Benchmark numbers, ablations and baseline numbers are extracted by AI from each paper's full text and can be wrong; the paper is the authority. Data: results index of 2026-11-02, citation counts from OpenAlex (2026-10-09).