开源权重模型能否替代GPT-4实现文化对齐机器翻译
原标题：Translating in style: An open-weight alternative to GPT-4 for culturally aligned machine translation
AI 摘要
该研究探讨用开源翻译模型 Hunyuan-MT-7B 替代 RASTA 框架中闭源 GPT-4 的可行性，评估基础模型、LoRA 监督微调与 GRPO 策略引导三个变体在正式度、礼貌度和亲密度三个风格维度及十二个语言对上的风格对齐表现。结果显示微调未带来统计显著的总体提升，开源模型整体不及 RASTA-GPT-4，差距集中在正式度维度，但在亲密度和中文方向上已接近甚至个别语言对超越 GPT-4。作者认为基于开源模型的 RASTA 流程在部分问题上已具备可行性，可提升可及性与可复现性。
正文节选
Translating in style: An open-weight alternative to GPT-4 for culturally aligned machine translation Abstract Machine translation systems often preserve meaning but systematically flatten linguistic style, a failure that is most acute when translating into non-Western languages. The RASTA framework of Havaldar et al. (2025) addresses this through retrieval-augmented stylistic alignment, but relies on GPT-4, a closed, proprietary model. We investigate whether an open-weight translation specialist