Research map / Multi-agent systems and games / Cooperative multi-agent RL

Adaptive Communication Range for Scalable Cooperative Multi-Agent Reinforcement Learning

NeurIPS 2026 · Multi-agent systems and games › Cooperative multi-agent RL · approach: Learned inter-agent communication

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes an adaptive communication range mechanism for scalable cooperative MARL that filters communication with low-relevance agents, aiming to avoid reliance on global state and fixed graph structure.

Problem

Cooperative multi-agent reinforcement learning scalability degrades as the number of agents grows; existing communication-filtering methods rely on global state and fixed graph distributions, limiting adaptability in large-scale settings.

Approach

Learns an adaptive communication range per agent so that communication is filtered by agent relevance rather than by global state or a fixed graph, improving adaptability at scale.

Closest papers by meaning

• Exponential Topology-enabled Scalable Communication in Multi-agent Reinforcement Learning (ICLR 2025 · similarity 0.91)
• Multi-Agent Reinforcement Learning with Communication-Constrained Priors (NeurIPS 2025 · similarity 0.91)
• Efficient Communication in Multi-Agent Reinforcement Learning with Implicit Consensus Generation (AAAI 2025 · similarity 0.89)
• Bridging Training and Execution via Dynamic Directed Graph-Based Communication in Cooperative Multi-Agent Systems (AAAI 2025 · similarity 0.89)
• Scalable Constrained Policy Optimization for Safe Multi-agent Reinforcement Learning (NeurIPS 2024 · similarity 0.89)
• HyperMARL: Adaptive Hypernetworks for Multi-Agent RL (NeurIPS 2025 · similarity 0.88)