Research map / Multi-agent systems and games / Cooperative multi-agent RL
Succinct and Robust Multi-Agent Communication With Temporal Message Control
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 |
|---|---|---|---|
| Use temporal message control with buffering and thresholded message transmission | QMIX agents · SMAC | average winning rate: 23% higher | Helped |
| Use temporal message control instead of VBC | SMAC hard scenarios · SMAC hard scenarios | communication overhead: 1.3× | Helped |
| Use temporal message control instead of SchedNet | SMAC hard scenarios · SMAC hard scenarios | communication overhead: 3.7× | Helped |
| Use temporal message control with message buffers | cooperative MARL agents · SMAC | communication overhead reduction: up to 80% | Helped |
| Use temporal message control rather than VBC | cooperative MARL agents · SMAC heavy loss | winning rate | Helped |
| Train AC using the light-loss pattern, then evaluate under heavy loss | AC · 3s_vs_5z | winning rate: 60% -> 2% | Hurt |
| Use TMC+VDN rather than the other algorithms | cooperative MARL agents · Predator-Prey | normalized reward: 1.24× higher | Helped |
| Use TMC+VDN rather than the other approaches | cooperative MARL agents · Predator-Prey | communication overhead: 3.2× reduction | Helped |
| Use TMC+VDN rather than the other algorithms | cooperative MARL agents · Cooperative Navigation | normalized reward: 1.35× higher | Helped |
| Use TMC+VDN rather than the other approaches | cooperative MARL agents · Cooperative Navigation | communication overhead: 2.9× reduction | Helped |
| Add the message-smoothing regularizer (ablation) | AC · 6h_vs_8z, medium loss | message distance below threshold within smoothing window: 93% | Helped |
| Remove both regularizers and the message-transmission limit (ablation) | TMC · SMAC | winning rate | Hurt |
Closest papers by meaning
- Multi-Agent Reinforcement Learning with Communication-Constrained Priors (NeurIPS 2025 · similarity 0.90)
- Exponential Topology-enabled Scalable Communication in Multi-agent Reinforcement Learning (ICLR 2025 · similarity 0.89)
- Efficient Communication in Multi-Agent Reinforcement Learning with Implicit Consensus Generation (AAAI 2025 · similarity 0.89)
- Robust Communicative Multi-Agent Reinforcement Learning with Active Defense (AAAI 2024 · similarity 0.89)
- Multi-Agent Coordination via Multi-Level Communication (NeurIPS 2024 · similarity 0.88)
- CoDe: Communication Delay-Tolerant Multi-Agent Collaboration via Dual Alignment of Intent and Timeliness (AAAI 2025 · similarity 0.87)
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.