Every Attention Matters: An Efficient Hybrid Architecture for Long-Context Reasoning

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Ling Team, Han, Bin, Tang, Caizhi, Liang, Chen, Zhang, Donghao, Yuan, Fan, Zhu, Feng, Gao, Jie, Hu, Jingyu, Li, Longfei, Li, Meng, Zhang, Mingyang, Jiang, Peijie, Jiao, Peng, Zhao, Qian, Yang, Qingyuan, Shen, Wenbo, Yang, Xinxing, Zhang, Yalin, Ren, Yankun, Zhao, Yao, Cao, Yibo, Sun, Yixuan, Zhang, Yue, Fang, Yuchen, Lin, Zibin, Cheng, Zixuan, Zhou, Jun
Format:	Preprint
Published:	2025
Subjects:	Machine Learning Artificial Intelligence 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/2510.19338