Rewrite to Translate, Translate to Reward: Reinforcement Learning for Source Rewriting in Machine Translation

[Submitted on 6 Jun 2026 (v1), last revised 10 Jun 2026 (this version, v2)]

View PDF HTML (experimental)

Abstract:Rewriting source text with large language models (LLMs) before translation has been shown to improve machine translation (MT) quality. However, we find that prompt-based rewriting can degrade translation quality rather than improve it, particularly when smaller LLMs, such as 4B-parameter models, are used. We argue that this limitation stems from the difficulty of controlling rewriting behavior through natural-language prompts alone: a rewrite is useful only if it improves downstream translation, yet existing prompt-based methods do not explicitly optimize for this signal. To address this issue, we propose RLSR (Reinforcement Learning for Source Rewriting), a reinforcement learning framework that trains the rewriting model with a reward based on the downstream translation-quality improvement produced by each rewrite. Experiments across six MT systems and 16 language pairs show that our 4B RLSR-trained rewriting models significantly outperform both the no-rewriting baseline and prompt-based rewriting baselines at the same model scale, while remaining competitive with baselines that use a 235B LLM.

Submission history

From: Boxuan Lyu [view email]
[v1] Sat, 6 Jun 2026 07:00:44 UTC (209 KB)
[v2] Wed, 10 Jun 2026 14:48:43 UTC (212 KB)