Answer First, Reason Later: Aligning Search Relevance via Mode-Balanced Reinforcement Learning

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Zhang, Shijie, Guo, Xiang, Guo, Rujun, Liu, Shaoyu, Wang, Xiaozhao, Jiang, Guanjun, Zhang, Kevin
Format:	Preprint
Published:	2026
Subjects:	Machine Learning
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Internet

https://arxiv.org/abs/2602.10006