构建基于语言感知指令微调的精准翻译定制大语言模型

昝畅通; 丁亮; 沈力; 詹忆冰; 杨兴浩; 刘伟锋

doi:10.1631/FITEE.2400458

Your Location：

Home >

Browse articles >

构建基于语言感知指令微调的精准翻译定制大语言模型

常规文章 | Updated：2025-09-04

- 构建基于语言感知指令微调的精准翻译定制大语言模型
  Enhanced Publication
- Building accurate translation-tailored large language models with language-aware instruction tuning
- “在自然语言处理领域，在提高机器翻译任务中大型语言模型（LLM）的准确性方面取得了突破。研究人员开发了一种两阶段微调算法，显著减少了脱靶翻译问题，提高了翻译质量。这项创新有效地解决了在不遵守指示的情况下以错误语言进行翻译的挑战。该方法涉及对翻译数据的LLM进行微调，然后引入额外的不可能性损失，以降低不正确翻译的概率。这一进步不仅提高了翻译的准确性，还保留了模型在其他任务上的性能。”
- 信息与电子工程前沿（英文） 2025年26卷第8期页码：1341-1355
- Affiliations：
  
  1.College of Control Science and Engineering, China University of Petroleum (East China), Qingdao 266580, China
  2.School of Computer Science, University of Sydney, New South Wales 2006, Australia
  3.JD Explore Academy, JD.com Inc., Beijing 100101, China
  4.School of Cyber Science and Technology, Shenzhen Campus of Sun Yat-sen University, Shenzhen 518107, China
- Author bio：
  
  ‡ Corresponding authors
- Funds：
  
  National Natural Science Foundation of China(62372468);Shandong Natural Science Foundation(ZR2023MF008);Major Basic Research Projects in Shandong Province(ZR2023ZD32);Qingdao Natural Science Foundation(23-2-1-161-zyyd-jch)
- DOI：10.1631/FITEE.2400458
  中图分类号： TP391
- 收稿：2024-03-30，
  
  修回：2024-11-27，
  
  纸质出版：2025-08
- Accepted：
Scan QR Code
昝畅通, 丁亮, 沈力, 等. 构建基于语言感知指令微调的精准翻译定制大语言模型[J]. 信息与电子工程前沿（英文）, 2025,26(8):1341-1355.

Changtong ZAN, Liang DING, Li SHEN, et al. Building accurate translation-tailored large language models with language-aware instruction tuning[J]. Frontiers of Information Technology & Electronic Engineering, 2025, 26(8): 1341-1355.
昝畅通, 丁亮, 沈力, 等. 构建基于语言感知指令微调的精准翻译定制大语言模型[J]. 信息与电子工程前沿（英文）, 2025,26(8):1341-1355. DOI： 10.1631/FITEE.2400458.

Changtong ZAN, Liang DING, Li SHEN, et al. Building accurate translation-tailored large language models with language-aware instruction tuning[J]. Frontiers of Information Technology & Electronic Engineering, 2025, 26(8): 1341-1355. DOI： 10.1631/FITEE.2400458.

浏览量

173

Downloads

CSCD

文章被引用时，请邮件提醒。

Submit

工具集

关联资源

GMCoT: a graph-augmented multimodal chain-of-thought reasoning framework for multi-label zero-shot learning

Mind the Gap: towards generalizable autonomous penetration testing via domain randomization and meta-reinforcement learning

Can large language models effectively process and execute financial trading instructions?

Four development stages of collective intelligence

构建基于语言感知指令微调的精准翻译定制大语言模型

Building accurate translation-tailored large language models with language-aware instruction tuning

DOI：10.1631/FITEE.2400458