Year: 2026
Type: conference-paper
Abstract: Reinforcement learning (RL) has shown strong promise for LLM-based machine translation, with recent methods such as GRPO demonstrating notable gains; nevertheless, translationoriented RL remains challenged by noisy learning signals arising from Monte Carlo return estimation, as well as a large trajectory space that favors global exploration over fine-grained local optimization.We introduce PEGRL, a twostage RL framework that uses post-editing as an auxiliary task to stabilize training and guide overall optimization.At each iteration, translation outputs are sampled to construct postediting inputs, allowing return estimation in the post-editing stage to benefit from conditioning on the current translation behavior, while jointly supporting both global exploration and fine-grained local optimization.A task-specific weighting scheme further balances the contributions of translation and post-editing objectives, yielding a biased yet more sample-efficient estimator.Experiments on English→Finnish, English→Turkish, and English↔Chinese show consistent gains over RL baselines, and for English→Turkish, performance on COMET-KIWI is comparable to advanced LLM-based systems (DeepSeek-V3.2).
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7166880276