一作ACL 2026Findings· Findings of ACL
PEGRL : Improving Machine Translation by Post-Editing Guided Reinforcement Learning
Yunzhi Shen · Hao Zhou · Xin Huang · Xue Han · Junlan Feng · Shujian Huang
摘要
Reinforcement learning (RL) has shown strong promise for LLM-based machine translation, with recent methods such as GRPO demonstrating notable gains; nevertheless, translation-oriented RL remains challenged by high-variance policy gradients induced by Monte Carlo baselines, as well as a large tra...