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 2302.07856cs.CLcs.LG
收藏

面向机器翻译的大语言模型短语级词典标注

Dictionary-based Phrase-level Prompting of Large Language Models for Machine Translation

Meta AI

Marjan Ghazvininejad, Hila Gonen, Luke Zettlemoyer

摘要:

大语言模型(LLM)通过提示展示了卓越的机器翻译(MT)能力,即使它们没有被明确训练用于此任务。然而,即使考虑到他们训练的数据量令人难以置信,LLM也很难翻译带有罕见单词的输入,这在低资源或域转移场景中很常见。我们表明,LLM提示可以提供一个有效的解决方案,以及罕见的话,通过使用双语词典的先验知识,提供控制提示。我们提出了一种新的方法,DiPMT,它提供了一组可能的翻译的一个子集的输入单词,从而使细粒度的短语级提示控制的LLM。大量的实验表明,DiPMT优于基线在低资源MT,以及域外MT。我们进一步提供了这种方法的优点和局限性的定性分析,包括实现的可控性的整体水平。

英文摘要:

Large language models (LLMs) demonstrate remarkable machine translation (MT) abilities via prompting, even though they were not explicitly trained for this task. However, even given the incredible quantities of data they are trained on, LLMs can struggle to translate inputs with rare words, which are common in low resource or domain transfer scenarios. We show that LLM prompting can provide an effective solution for rare words as well, by using prior knowledge from bilingual dictionaries to provide control hints in the prompts. We propose a novel method, DiPMT, that provides a set of possible translations for a subset of the input words, thereby enabling fine-grained phrase-level prompted control of the LLM. Extensive experiments show that DiPMT outperforms the baseline both in low-resource MT, as well as for out-of-domain MT. We further provide a qualitative analysis of the benefits and limitations of this approach, including the overall level of controllability that is achieved.

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。

↑