IntGrad MT: Eliciting LLMs' Machine Translation Capabilities with Sentence Interpolation and Gradual MT

Recent Large Language Models (LLMs) have demonstrated strong performance in translation without needing to be finetuned on additional parallel corpora. However, they still underperform for low-resource language pairs. Previous works have focused on mitigating this issue by leveraging relevant few-shot examples or external resources such as dictionaries or grammar books, making models heavily reliant on these nonparametric sources of information. In this paper, we propose a novel method named IntGrad MT that focuses on fully exploiting an LLM's inherent translation capability. IntGrad MT achieves this by constructing a chain of few-shot examples, each consisting of a source sentence and the model's own translation, that rise incrementally in difficulty. IntGrad MT employs two techniques: Sentence Interpolation, which generates a sequence of sentences that gradually change from an easy sentence to translate to a difficult one, and Gradual MT, which sequentially translates this chain using translations of earlier sentences as few-shot examples for the translation of subsequent ones. With this approach, we observe a substantial enhancement in the xCOMET scores of various LLMs for multiple languages, especially in low-resource languages such as Hindi(8.26), Our approach presents a practical way of enhancing LLMs' performance without extra training. Recent Large Language Models (LLMs) have shown strong performance in translation tasks without the need for fine-tuning on specific parallel datasets. Previous studies have demonstrated that LLMs' translation capabilities are reliable in most use cases, particularly when the source and target language are high-resource languages (Zhu et al., 2024; Robinson et al., 2023; Jiao et al., 2023). However, because LLMs require training on large corpora, they still face challenges when translating low-resource languages that are not sufficiently represented in the training corpora.(Stap Previous research has attempted to address these challenges by leveraging the in-context-learning capabilities of large language models (LLMs), particularly through the use of external knowledge such as few-shot examples or dictionaries during inference.