On-Policy Distillation (OPD) facilitates the transfer of knowledge from domain expert to student in the post-training phase of Large Language Models (LLMs). …
机构:港科大(广州)
来源:arXiv 2609.08341 | AI4Papers 论文推荐平台