@InProceedings{he-EtAl:2026:wmt,
  author    = {He, Yu  and  Lan, Xiaoqing  and  Wei, Daimeng  and  GUO, Jiaxin  and  Luo, Yuanchang  and  Shang, Hengchao  and  Li, Zongyao  and  Yang, Jinlong  and  Wu, Zhanglin  and  Huang, Boqi},
  title     = {HW-TSC's Submission to the WMT 2026 General Machine Translation Track},
  booktitle      = {Proceedings of the Eleventh Conference on Machine Translation},
  month          = {October},
  year           = {2026},
  address        = {Budapest, Hungary},
  publisher      = {Association for Computational Linguistics},
  pages     = {1265--1270},
  abstract  = {Machine translation based on large language models (LLMs) has achieved remarkable progress in recent years. However, relying on a single translation model and fixed translation instructions often limits translation quality, especially for multilingual and multi-domain scenarios. To address this issue, we propose HW-TSC-Agent, a multi-stage translation agent framework that integrates translation model fine-tuning, dual-model candidate generation, and instruction-aware translation refinement into a unified pipeline. First, we perform Supervised Fine-Tuning (SFT) and Contrastive Preference Optimization (CPO) on openPangu-Embedded-7B using high-quality bilingual data to enhance its multilingual translation capabilities. The fine-tuned openPangu-Embedded-7B and HY-MT2-7B are then used to generate candidate translations. Finally, source-language analysis, candidate translations, and task-specific instructions are incorporated into a dynamic prompt, which guides DeepSeek-V4-Flash to compare, refine, and fuse the candidate translations to produce the final output. Our system is submitted to the WMT2026 General Machine Translation Shared Task, covering 13 translation directions. On the WMT2025 English-to-Chinese development test set, enriching the dynamic prompt generally improves translation quality over an instruction-only DeepSeek-V4-Flash baseline.},
  url       = {https://aclanthology.org/2026.wmt-1.66}
}