One-Token Rollout: Guiding Supervised Fine-Tuning of LLMs with Policy Gradient

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Ming, Rui, Wu, Haoyuan, Hu, Shoubo, He, Zhuolun, Yu, Bei
Format:	Preprint
Published:	2025
Subjects:	Computation and Language
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Internet

https://arxiv.org/abs/2509.26313