Distributionally Robust Cooperative Multi-agent Reinforcement Learning with Value Factorization

ICLR 2026 Conference Submission9907 Authors

Keywords: distributionally robust RL, cooperative multi-agent RL; Centralized training decentralized execution

TL;DR: We develop distributionally robust cooperative MARL algorithms based on a novel value factorization principle that accounts for environmental uncertainty.

Abstract: Cooperative multi-agent reinforcement learning (MARL) commonly adopts centralized training with decentralized execution, where value-factorization methods enforce the individual-global-maximum (IGM) principle so that decentralized greedy actions recover the team-optimal joint action. However, the reliability of this recipe in real-world settings remains uncertain due to environmental uncertainties arising from the sim-to-real gap, model mismatch, system noise. We address this gap by introducing Distributionally robust IGM (DrIGM), a principle that requires each agent's robust greedy action to align with the robust team-optimal joint action. We show that DrIGM holds for a novel definition of robust individual action values, which is compatible with decentralized greedy execution and yields a provable robustness guarantee for the whole system. Building on this foundation, we derive DrIGM-compliant robust variants of existing value-factorization architectures (e.g., VDN/QMIX/QTRAN) that (i) train on robust Q-targets, (ii) preserve scalability, and (iii) integrate seamlessly with existing codebases without bespoke per-agent reward shaping. Empirically, on high-fidelity SustainGym simulators, our methods consistently improve out-of-distribution performances.

Primary Area: reinforcement learning

Submission Number: 9907