Learning-From-Mistakes Prompting for Indigenous Language Translation

TL;DR

This paper explores novel prompting techniques leveraging large language models to improve translation accuracy for extremely low-resource indigenous languages, using minimal data and in-context learning strategies.

Contribution

Introduces Learning-from-Mistakes Prompting and other methods to enhance low-resource language translation with LLMs, addressing data scarcity challenges.

Findings

LLMs can effectively translate low-resource languages with proper prompting.

Learning-from-Mistakes Prompting improves translation accuracy.

Techniques outperform baseline methods on limited data sets.

Abstract

Using large language models, this paper presents techniques to improve extremely low-resourced indigenous language translations. Our approaches are grounded in the use of (1) the presence of a datastore consisting of a limited number of parallel translation examples, (2) the inherent capabilities of LLMs like GPT-3.5, and (3) a word-level translation dictionary. We harness the potential of LLMs and in-context learning techniques in such a setting for using LLMs as universal translators for extremely low-resourced languages. Our methodology hinges on utilizing LLMs as language compilers for selected language pairs, hypothesizing that they could internalize syntactic structures to facilitate accurate translation. We introduce three techniques: KNNPrompting with Retrieved Prompting Context, Chain-of-Thought Prompting and Learningfrom-Mistakes Prompting, with the last method addressing past errors. The evaluation results suggest that, even with limited corpora, LLMs can effectively translate extremely low-resource languages when paired with proper prompting.