Rohan Paul@rohanpaul_aiMuon优化器几乎使AI智能体的成功率翻倍，表明其强化学习价值取决于信用分配和学习率。 通过GiGPO（比较重复状态下的动作），Muon将后期成功率从0.29提升至0.55。 优化器不能单独评判，因为周围的强化学习设置可能决定其成败。 – arxiv.org/abs/2607.16169
Muon optimizer nearly doubled an AI agent's success， showing its reinforcement-learning value depends on credit assignment and learning rate.
With GiGPO， which compares actions from repeated states， Muon raised late success from 0.29 to 0.55.
An optimizer cannot be judged alone because the surrounding reinforcement-learning setup may decide whether it helps or fails.
Rohan Paul@rohanpaul_ai · XMuon优化器几乎使AI智能体的成功率翻倍，表明其强化学习价值取决于信用分配和学习率。 通过GiGPO（比较重复状态下的动作），Muon将后期成功率从0.29提升至0.55。 优化器不能单独评判，因为周围的强化学习设置可能决定其成败。 – arxiv.org/abs/2607.16169
Muon optimizer nearly doubled an AI agent's success， showing its reinforcement-learning value depends on credit assignment and learning rate.
With GiGPO， which compares actions from repeated states， Muon raised late success from 0.29 to 0.55.
An optimizer cannot be judged alone because the surrounding reinforcement-learning setup may decide whether it helps or fails.