Reproduce “Multi-Agent MDP Homomorphic Networks” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Elise van der Pol, Herke van Hoof, Frans A. Oliehoek, Max Welling
The authors introduce an equivariant message-passing network for cooperative agents that coordinate using local observations and neighbor messages. It factors global rotations into compatible changes to local observations, interactions, and agent arrangements, allowing policies to execute locally. In tested traffic and wildlife tasks, it improves learning over distributed standard and rotation-augmentation baselines; evidence covers exact discrete rotations and discrete actions.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.