@inproceedings{wang-etal-2026-m2po,
    title = "{M}$^2${PO}: Multi-Perspective Multi-Pair Preference Optimization for Machine Translation",
    author = "Wang, Hao  and
      Xu, Linlong  and
      Liu, Heng  and
      Liu, Yangyang  and
      Zhao, Xiaohu  and
      Zeng, Bo  and
      Shao, Liangying  and
      Dong, Yichen  and
      Wu, Xinwei  and
      Zhou, Jiang  and
      Dong, Tianyu  and
      Zeng, Xiangxiang  and
      Wang, Longyue  and
      Luo, Weihua",
    editor = "Liakata, Maria  and
      Moreira, Viviane P.  and
      Zhang, Jiajun  and
      Jurgens, David",
    booktitle = "Proceedings of the 64th Annual Meeting of the {A}ssociation for {C}omputational {L}inguistics (Volume 1: Long Papers)",
    month = jul,
    year = "2026",
    address = "San Diego, California, United States",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-acl/2026.acl-long.469/",
    pages = "10315--10336",
    ISBN = "979-8-89176-390-6",
    abstract = "Aligning Large Language Models (LLMs) to human preferences is pivotal for Machine Translation (MT), yet current approaches are often hindered by misleading reward signals. Our analysis reveals that prevailing Quality Estimation (QE) models exhibit a systematic blind spot towards **partial errors**{---}specifically partial hallucinations and omissions{---}often favoring superficially fluent but unfaithful translations. To address this, we propose **M$^2$PO** (**M**ulti-Perspective **M**ulti-Pair **P**reference **O**ptimization), a data-centric framework for preference optimization in machine translation. First, to correct the bias towards fluency, M$^2$PO uses a multi-perspective alignment mechanism that decouples semantic fidelity from fluency, prioritizing faithfulness via a curriculum strategy. Second, with the bias corrected, partial errors fall between perfect and severely incorrect translations, making them inefficient to learn via standard best-versus-worst comparisons. We thus introduce a multi-pair objective that leverages the full candidate list to capture these fine-grained error signals. Experiments on WMT23, WMT24, and FLORES-200 show that M$^2$PO enables a 9B model to outperform leading open-source baselines and achieve parity with proprietary models like GPT-4o and Gemini-2.0-Flash, demonstrating significant potential for efficient, high-fidelity LLM-based translation."
}