A Unified View of Attention and Residual Sinks: Outlier-Driven Rescaling is Essential for Transformer Training

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Qiu, Zihan, Huang, Zeyu, Wen, Kaiyue, Jin, Peng, Zheng, Bo, Zhou, Yuxin, Huang, Haofeng, Wang, Zekun, Li, Xiao, Zhang, Huaqing, Xu, Yang, Lian, Haoran, Zhang, Siqi, Men, Rui, Zhang, Jianwei, Titov, Ivan, Liu, Dayiheng, Zhou, Jingren, Lin, Junyang
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!