@inproceedings{chen-etal-2025-realign,
    title = "{R}e{A}lign: Structured Revision for Small Language Model Alignment",
    author = "Chen, Ruijun  and
      Guo, Jiajian  and
      Chen, Hongzhan  and
      Wan, Fanqi  and
      Wang, Qifan  and
      Quan, Xiaojun",
    editor = "Christodoulopoulos, Christos  and
      Chakraborty, Tanmoy  and
      Rose, Carolyn  and
      Peng, Violet",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2025",
    month = nov,
    year = "2025",
    address = "Suzhou, China",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/author-page-yu-wang-polytechnic/2025.findings-emnlp.642/",
    doi = "10.18653/v1/2025.findings-emnlp.642",
    pages = "12005--12020",
    ISBN = "979-8-89176-335-7",
    abstract = "Aligning small language models with human preferences is challenging, as weak policies struggle to generate informative on-policy samples and suffer from unstable gradients when trained on off-policy signals from stronger models. In this work, we propose ReAlign, a training framework that combines the stability of on-policy learning with the guidance of reviser-assisted supervision. In the ReAlign, we first train a lightweight reviser to improve policy-generated responses using preference-based supervision, conditioned on both the prompt and the initial output. And then, the policy is optimized using a combination of standard on-policy preference pairs and reviser-enhanced pairs constructed as a structured revision task, where the latter provide richer, more learnable feedback. Experimental results on AlpacaEval-2 and Arena-Hard demonstrate that ReAlign significantly boosts alignment performance for weak policies, outperforming strong preference optimization baselines."
}