STAPO: Stabilizing Reinforcement Learning for LLMs by Silencing Rare Spurious Tokens

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Liu, Shiqi, He, Zeyu, Zhan, Guojian, Tao, Letian, Zheng, Zhilong, Wu, Jiang, Wang, Yinuo, Guan, Yang, Sheng, Kehua, Zhang, Bo, Li, Keqiang, Duan, Jingliang, Li, Shengbo Eben
Format:	Preprint
Published:	2026
Subjects:	Computation and Language Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Internet

https://arxiv.org/abs/2602.15620