@InProceedings{wang-EtAl:2026:wmt3,
  author    = {Wang, Zhenhan  and  Fan, Fengzhao  and  Huang, Yuxin  and  Tan, Kaiwen  and  Yu, Zhengtao  and  Gao, Shengxiang  and  Mao, Cunli  and  Zhang, Siqi  and  Jiang, Shuting},
  title     = {Xiaoyu-MT: Kunming University of Science and Technology at WMT 2026 Chinese-Southeast Asian Multilingual MT Task},
  booktitle      = {Proceedings of the Eleventh Conference on Machine Translation},
  month          = {October},
  year           = {2026},
  address        = {Budapest, Hungary},
  publisher      = {Association for Computational Linguistics},
  pages     = {2378--2387},
  abstract  = {This paper presents Xiaoyu-MT, our submission system for the WMT 2026 Chinese-Southeast Asian Multilingual MT Task, focusing on bidirectional translation between Chinese and seven Southeast Asian languages. To address limited parallel resources and imbalanced multilingual data distributions, we select Gemma-4-12B-it as the base model and construct multilingual training data by combining official WMT resources with additional monolingual and bilingual corpora. Xiaoyu-MT adopts a multi-stage training pipeline consisting of Continual Pre-Training (CPT), Supervised Fine-Tuning (SFT), and Direct Preference Optimization (DPO). During SFT, we introduce a language-aware Prompt encoding mechanism to explicitly model source language, target language, and their directed translation relationships through structured Prompt representations and Prompt compression. DPO further optimizes translation preferences using preference data constructed from reference and model-generated translations. Official evaluation results on WMT 2026 show that Xiaoyu-MT achieves an overall BLEU score of 34.73, a COMET score of 81.87, and a Final Result score of 61.80, demonstrating its effectiveness for multilingual and low-resource machine translation scenarios.},
  url       = {https://aclanthology.org/2026.wmt-1.181}
}