Diagonal-Tiled Mixed-Precision Attention for Efficient Low-Bit MXFP Inference

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Ding, Yifu, Zhang, Xinhao, Guo, Jinyang
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!