@article{howoffpolicycangrpobemugrpoforefficientl,
  title = {How Off-Policy Can GRPO Be? Mu-GRPO for Efficient LLM Reinforcement Learning},
  author = {Minghao Tian and Yunfei Xie and Chen Wei},
  year = {2026},
  eprint = {2605.17570},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2605.17570},
}