Tag
This paper investigates the use of the Muon optimizer in reinforcement learning post-training, finding that applying Muon to hidden weight matrices significantly improves success rates on ALFRED tasks compared to AdamW, with results dependent on the advantage estimator and learning rate.