Prompting Large Language Model for Machine Translation: A Case Study
This work addresses the under-explored problem of prompting for machine translation, offering insights for researchers and practitioners, but it is incremental as it builds on existing prompting research without introducing a new paradigm.
The study systematically investigates prompting strategies for machine translation using the GLM-130B model, finding that factors like the number and quality of prompt examples affect performance, with pseudo parallel examples from monolingual data improving translation and transfer learning across settings showing potential gains.
Research on prompting has shown excellent performance with little or even no supervised training across many tasks. However, prompting for machine translation is still under-explored in the literature. We fill this gap by offering a systematic study on prompting strategies for translation, examining various factors for prompt template and demonstration example selection. We further explore the use of monolingual data and the feasibility of cross-lingual, cross-domain, and sentence-to-document transfer learning in prompting. Extensive experiments with GLM-130B (Zeng et al., 2022) as the testbed show that 1) the number and the quality of prompt examples matter, where using suboptimal examples degenerates translation; 2) several features of prompt examples, such as semantic similarity, show significant Spearman correlation with their prompting performance; yet, none of the correlations are strong enough; 3) using pseudo parallel prompt examples constructed from monolingual data via zero-shot prompting could improve translation; and 4) improved performance is achievable by transferring knowledge from prompt examples selected in other settings. We finally provide an analysis on the model outputs and discuss several problems that prompting still suffers from.