Beyond GRPO and On-Policy Distillation: An Empirical Sparse-to-Dense Reward Principle for Language-Model Post-Training

Yuanda Xu, Hejian Sang, Zhengze Zhou, Ran He, Zhipeng Wang, Alborz GeramifardarXiv 2026 · DOI

BibTeX
@misc{xu2026beyond,
  author = {Yuanda Xu and Hejian Sang and Zhengze Zhou and Ran He and Zhipeng Wang and Alborz Geramifard},
  title = {{Beyond GRPO and On-Policy Distillation: An Empirical Sparse-to-Dense Reward Principle for Language-Model Post-Training}},
  year = {2026},
  eprint = {2605.12483},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2605.12483},
}

Loading related content...

Figure