Leveraging GPT-4 for Automatic Translation Post-Editing

While Neural Machine Translation (NMT) represents the leading approach to Machine Translation (MT), the outputs of NMT models still require translation post-editing to rectify errors and enhance quality under critical settings.In this work, we formalize the task of direct translation post-editing with Large Language Models (LLMs) and explore the use of GPT-4 to automatically post-edit NMT outputs across several language pairs.Our results demonstrate that GPT-4 is adept at translation post-editing, producing meaningful and trustworthy edits to translations that help improve its general quality as well as remove different classes of major errors in translations.In particular, human evaluations on assessing edit trustworthiness show that GPT-4 exhibits a large improvement over the prior state-of-the-art LLM.Notably, we improve upon state-of-the-art performance on WMT-22 English-Chinese, English-German, Chinese-English and German-English language pairs using GPT-4 based post-editing, as evaluated by state-of-the-art MT quality metrics.However, we also show that GPT-4 could produce hallucinated edits, thereby urging caution in its use as an expert translation post-editor. LimitationsWe proposed a formalization to study direct automatic post-editing with state-of-the-art LLMs and investigated a number of research questions through this formalization.However, we only have API-level access to GPT-4.Even though we conducted our experiments on WMT-22 test sets and system outputs, the curation of which falls outside the cut-off date for GPT-4 training data; due to only black-box access to the model, we cannot rule out the possibility of data contamination, even on the WMT-22 test sets.