arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2402.16379cs.CLcs.AI
收藏

TEaR: 借助系统性自我反思改进基于大语言模型的机器翻译

TEaR: Improving LLM-based Machine Translation with Systematic Self-Refinement

Zhejiang University(浙江大学);Tencent(腾讯);Angelalign Technology Inc.(时代天使科技有限公司)

Zhaopeng Feng, Yan Zhang, Hao Li, Bei Wu, Jiayu Liao, Wenqiang Liu, Jun Lang, Yang Feng, Jian Wu, Zuozhu Liu

AI总结:

针对大语言模型机器翻译仍存错误的问题,提出名为TEaR的翻译-评估-反思系统性自我修正框架,可提升多场景翻译质量,兼具强系统性与可解释性,并探究了通用LLM翻译与评估能力的关联。

摘要:

大语言模型(LLM)已在机器翻译(MT)领域取得了令人瞩目的成果。然而,人工细致评估显示,LLM生成的译文仍存在多种错误。值得注意的是,将这类错误信息反馈给LLM可触发自我反思,进而提升翻译性能。受这些发现启发,我们提出了一个基于LLM的系统性自我反思翻译框架,名为**TEaR**,全称为**T**ranslate(翻译)、**E**stimate(评估)、a**n**d **R**efine(反思),是该方向的一项重要进展。我们的研究结果表明:1)该自我反思框架可有效帮助LLM提升多语言翻译质量,覆盖从高资源语言到低资源语言的翻译场景,无论是以英语为中心还是以其他语言为中心的翻译方向均适用;2)TEaR具备更强的系统性和可解释性;3)不同的评估策略会产生不同影响,直接决定最终修正的效果。此外,传统神经翻译模型与评估模型相互独立,受能力限制通常仅聚焦单一任务,而通用LLM具备同时承担两项任务的能力。我们还开展了跨模型修正实验,以探究通用LLM翻译能力与评估能力之间的潜在关联。代码与数据可在https://github.com/fzp0424/self_correct_mt 获取。

英文摘要:

Large Language Models (LLMs) have achieved impressive results in Machine Translation (MT). However, careful evaluations by human reveal that the translations produced by LLMs still contain multiple errors. Importantly, feeding back such error information into the LLMs can lead to self-refinement and result in improved translation performance. Motivated by these insights, we introduce a systematic LLM-based self-refinement translation framework, named \textbf{TEaR}, which stands for \textbf{T}ranslate, \textbf{E}stimate, \textbf{a}nd \textbf{R}efine, marking a significant step forward in this direction. Our findings demonstrate that 1) our self-refinement framework successfully assists LLMs in improving their translation quality across a wide range of languages, whether it's from high-resource languages to low-resource ones or whether it's English-centric or centered around other languages; 2) TEaR exhibits superior systematicity and interpretability; 3) different estimation strategies yield varied impacts, directly affecting the effectiveness of the final corrections. Additionally, traditional neural translation models and evaluation models operate separately, often focusing on singular tasks due to their limited capabilities, while general-purpose LLMs possess the capability to undertake both tasks simultaneously. We further conduct cross-model correction experiments to investigate the potential relationship between the translation and evaluation capabilities of general-purpose LLMs. Our code and data are available at https://github.com/fzp0424/self_correct_mt

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑