xTower: A Multilingual LLM for Explaining and Correcting Translation Errors

TL;DR

xTower is a multilingual large language model designed to explain translation errors and generate corrected translations, improving overall translation quality and aiding user understanding.

Contribution

This paper introduces xTower, a novel LLM that provides explanations for translation errors and generates corrected translations, enhancing interpretability and translation accuracy.

Findings

xTower produces plausible, helpful explanations of translation errors

xTower significantly improves translation quality in various setups

Expert evaluators find explanations related and useful

Abstract

While machine translation (MT) systems are achieving increasingly strong performance on benchmarks, they often produce translations with errors and anomalies. Understanding these errors can potentially help improve the translation quality and user experience. This paper introduces xTower, an open large language model (LLM) built on top of TowerBase designed to provide free-text explanations for translation errors in order to guide the generation of a corrected translation. The quality of the generated explanations by xTower are assessed via both intrinsic and extrinsic evaluation. We ask expert translators to evaluate the quality of the explanations across two dimensions: relatedness towards the error span being explained and helpfulness in error understanding and improving translation quality. Extrinsically, we test xTower across various experimental setups in generating translation corrections, demonstrating significant improvements in translation quality. Our findings highlight xTower's potential towards not only producing plausible and helpful explanations of automatic translations, but also leveraging them to suggest corrected translations.