RWARE

RWARE is a benchmark used to evaluate cooperative multi-agent reinforcement learning algorithms, assessing their performance in scenarios involving multiple agents working together.

Papers using RWARE (3)

Centralized Permutation Equivariant Policy for Cooperative Multi-Agent Reinforcement Learning2025Enhancing Cooperative Multi-Agent Reinforcement Learning with State Modelling and Adversarial Exploration2025Oryx: a Scalable Sequence Model for Many-Agent Coordination in Offline MARL2025