Iterative Translation Refinement with Large Language Models

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Chen, Pinzhen, Guo, Zhicheng, Haddow, Barry, Heafield, Kenneth
Natura: Preprint
Pubblicazione: 2023
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866909186807300096
author Chen, Pinzhen
Guo, Zhicheng
Haddow, Barry
Heafield, Kenneth
author_facet Chen, Pinzhen
Guo, Zhicheng
Haddow, Barry
Heafield, Kenneth
contents We propose iteratively prompting a large language model to self-correct a translation, with inspiration from their strong language understanding and translation capability as well as a human-like translation approach. Interestingly, multi-turn querying reduces the output's string-based metric scores, but neural metrics suggest comparable or improved quality. Human evaluations indicate better fluency and naturalness compared to initial translations and even human references, all while maintaining quality. Ablation studies underscore the importance of anchoring the refinement to the source and a reasonable seed translation for quality considerations. We also discuss the challenges in evaluation and relation to human performance and translationese.
format Preprint
id arxiv_https___arxiv_org_abs_2306_03856
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Iterative Translation Refinement with Large Language Models
Chen, Pinzhen
Guo, Zhicheng
Haddow, Barry
Heafield, Kenneth
Computation and Language
Artificial Intelligence
We propose iteratively prompting a large language model to self-correct a translation, with inspiration from their strong language understanding and translation capability as well as a human-like translation approach. Interestingly, multi-turn querying reduces the output's string-based metric scores, but neural metrics suggest comparable or improved quality. Human evaluations indicate better fluency and naturalness compared to initial translations and even human references, all while maintaining quality. Ablation studies underscore the importance of anchoring the refinement to the source and a reasonable seed translation for quality considerations. We also discuss the challenges in evaluation and relation to human performance and translationese.
title Iterative Translation Refinement with Large Language Models
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2306.03856