Enhancing Paraphrase Type Generation: The Impact of DPO and RLHF Evaluated with Human-Ranked Data 文章

ArXiv CS.CL2026-06-03NEWSen作者: Christopher Lee L\"ubbers
查看原文 →
关系图谱

Enhancing Paraphrase Type Generation: The Impact of DPO and RLHF Evaluated with Human-Ranked Data · 相关技术

相关技术

Direct Preference Optimization