Research map / Multi-agent systems and games / Cooperative multi-agent RL
Divergence-Regularized Multi-Agent Actor-Critic
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 DMAC to COMA | COMA · Randomly generated stochastic game | episode reward | Helped |
| Add DMAC to MAAC | MAAC · SMAC five tasks | win rate or episode reward | Helped |
| Add DMAC to QMIX | QMIX · SMAC tasks | win rate | Helped |
| Add DMAC to DOP | DOP · SMAC tasks | win rate or episode reward | Helped |
| Add DMAC to COMA | COMA · SMAC tasks | win rate | Helped |
| Add DMAC to FOP | FOP · SMAC 3s_vs_3z | convergence speed | Helped |
| Add DMAC to FOP | FOP · SMAC 2s3z, 3s5z, 2c_vs_64zg, and MMM2 | win rate | Helped |
| Use DMAC with omega=0.2 | COMA · Randomly generated stochastic game | cover rate | Helped |
| Increase the regularization coefficient to omega=10 or omega=100 (ablation) | COMA+DMAC · Randomly generated stochastic game | episode reward | Hurt |
| Decrease the regularization coefficient to omega=0.001 or omega=0.01 (ablation) | COMA+DMAC · Randomly generated stochastic game | episode reward | Hurt |
| Use a moderate regularization coefficient, omega=0.1, omega=0.2, or omega=1 (ablation) | COMA+DMAC · Randomly generated stochastic game | episode reward | Helped |
| Use DAPO with off-policy correction | DAPO · SMAC 3m | performance | Hurt |
Closest papers by meaning
- ComaDICE: Offline Cooperative Multi-Agent Reinforcement Learning with Stationary Distribution Shift Regularization (ICLR 2025 · similarity 0.89)
- Maximum Entropy Heterogeneous-Agent Reinforcement Learning (ICLR 2024 · similarity 0.89)
- Measuring Mutual Policy Divergence for Multi-Agent Sequential Exploration (NeurIPS 2024 · similarity 0.89)
- Convergence Theorems for Entropy-Regularized and Distributional Reinforcement Learning (NeurIPS 2025 · similarity 0.89)
- Optimistic Multi-Agent Policy Gradient (ICML 2024 · similarity 0.88)
- Cooperative Policy Agreement: Learning Diverse Policy for Offline MARL (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.