Research map / Multi-agent systems and games / Cooperative multi-agent RL
Randomized Entity-wise Factorization for Multi-Agent Reinforcement Learning
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Add randomized entity-wise factorization as an auxiliary objective to QMIX (Attention) | QMIX (Attention) · 3-8sz, 3-8csz, 3-8MMM | win rate | Helped |
| Add randomized entity-wise factorization as an auxiliary objective to QMIX (UPDeT) | QMIX (UPDeT) · 3-8sz, 3-8MMM | win rate | Helped |
| Add randomized entity-wise factorization using random subgroups rather than oracle-defined groups (ablation) | REFIL · Group Matching Game | win rate | Helped |
| Use fixed oracle groups instead of randomly sampled subgroups (ablation) | REFIL · Group Matching Game | win rate | Hurt |
| Restrict random subgroup sampling to known relevant entities (ablation) | REFIL · Group Matching Game | win rate | No clear effect |
| Remove the auxiliary imagined-learning loss (ablation) | REFIL · 3-8sz, 3-8csz, 3-8MMM | win rate | Hurt |
| Replace the nonlinear monotonic mixing network with summation (ablation) | REFIL (VDN) · 3-8sz, 3-8csz, 3-8MMM | win rate | Hurt |
| Replace attention with mean pooling for entity aggregation (ablation) | QMIX (Attention) · 3-8sz, 3-8csz, 3-8MMM | win rate | Hurt |
| Replace attention with max pooling for entity aggregation (ablation) | QMIX (Attention) · 3-8sz, 3-8csz, 3-8MMM | win rate | Hurt |
| Train exclusively with random factorization instead of combining it with the standard QMIX loss (ablation) | REFIL · 3-8sz | win rate | Hurt |
| Apply random factorization to UPDeT (ablation) | QMIX (UPDeT) · 3-8MMM | time steps to win: 22% fewer time steps | Helped |
Closest papers by meaning
- Towards Understanding Cooperative Multi-Agent Q-Learning with Value Factorization (NeurIPS 2021 · similarity 0.88)
- ConcaveQ: Non-monotonic Value Function Factorization via Concave Representations in Deep Multi-Agent Reinforcement Learning (AAAI 2024 · similarity 0.86)
- Imagine, Initialize, and Explore: An Effective Exploration Method in Multi-Agent Reinforcement Learning (AAAI 2024 · similarity 0.85)
- REValueD: Regularised Ensemble Value-Decomposition for Factorisable Markov Decision Processes (ICLR 2024 · similarity 0.85)
- Towards Principled Unsupervised Multi-Agent Reinforcement Learning (NeurIPS 2025 · similarity 0.85)
- FactorSim: Generative Simulation via Factorized Representation (NeurIPS 2024 · 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.