作者位次 #5/8ICLR 2026Poster· Reinforcement Learning / Everything Else
Balancing the Experts: Unlocking LoRA-MoE for GRPO via Mechanism-Aware Rewards
Changlian Ma · Zizheng Huang · Xiangyu Zeng · Yi Wang · Cheng Liang · Kun Tian · Xinhai Zhao · Limin Wang