Hypermarl: Adaptive Hypernetworks For Multi-agent RL

Abstract

Adaptive cooperation in multi-agent reinforcement learning (MARL) requires policies to express homogeneous, specialised, or mixed behaviours, yet achieving this adaptivity remains a critical challenge. While parameter sharing (PS) is standard for efficient learning, it notoriously suppresses the behavioural diversity required for specialisation. This failure is largely due to cross-agent gradient interference, a problem we find is surprisingly exacerbated by the common practice of coupling agent IDs with observations. Existing remedies typically add complexity through altered objectives, manual preset diversity levels, or sequential updates -- raising a fundamental question: can shared policies adapt without these intricacies? We propose a solution built on a key insight: an agent-conditioned hypernetwork can generate agent-specific parameters and decouple observation- and agent-conditioned gradients, directly countering the interference from coupling agent IDs with observations. Our r

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keytessera2024hypermarl

Related papers

• Revisiting Some Common Practices In Cooperative Multi-agent Reinforcement Learning (2022)0.00
• Adaptability In Multi-agent Reinforcement Learning: A Framework And Unified Review (2025)0.00
• Heterogeneous Multi-agent Reinforcement Learning For Zero-shot Scalable Collaboration (2024)6.34
• Cooperative Multi-agent RL With Communication Constraints (2026)0.00
• Multi-agent Reinforcement Learning In Stochastic Networked Systems (2020)0.00
• Individual Specialization In Multi-task Environments With Multiagent Reinforcement Learners (2019)0.00
• Enhancing Heterogeneous Multi-agent Cooperation In Decentralized MARL Via Gnn-driven Intrinsic Rewards (2024)0.00
• Low-rank Agent-specific Adaptation (lorasa) For Multi-agent Policy Learning (2025)0.00