Shortcomings of LLMs for Low-Resource Translation: Retrieval and Understanding are Both the Problem

TL;DR

This paper explores the capabilities and limitations of large language models in low-resource language translation, focusing on retrieval methods and understanding, with experiments on Southern Quechua to Spanish translation.

Contribution

It provides an empirical analysis of how different context types and retrieval methods affect LLM performance in low-resource translation tasks.

Findings

Small LLMs can utilize prompt context for zero-shot translation with minimal linguistic info

Context type and retrieval method significantly influence translation quality

Limitations exist in applying LLMs to most of the world's low-resource languages

Abstract

This work investigates the in-context learning abilities of pretrained large language models (LLMs) when instructed to translate text from a low-resource language into a high-resource language as part of an automated machine translation pipeline. We conduct a set of experiments translating Southern Quechua to Spanish and examine the informativity of various types of context retrieved from a constrained database of digitized pedagogical materials (dictionaries and grammar lessons) and parallel corpora. Using both automatic and human evaluation of model output, we conduct ablation studies that manipulate (1) context type (morpheme translations, grammar descriptions, and corpus examples), (2) retrieval methods (automated vs. manual), and (3) model type. Our results suggest that even relatively small LLMs are capable of utilizing prompt context for zero-shot low-resource translation when provided a minimally sufficient amount of relevant linguistic information. However, the variable effects of context type, retrieval method, model type, and language-specific factors highlight the limitations of using even the best LLMs as translation systems for the majority of the world's 7,000+ languages and their speakers.