SeVeDo: A Heterogeneous Transformer Accelerator for Low-Bit Inference via Hierarchical Group Quantization and SVD-Guided Mixed Precision

SeVeDo: A Heterogeneous Transformer Accelerator for Low-Bit Inference via Hierarchical Group Quantization and SVD-Guided Mixed Precision

15 December 2025

Yuseon Choi

Sangjin Kim

Jungjun Oh

Byeongcheol Kim

Hoi-Jun Yoo

ArXiv (abs)PDF HTML

Papers citing "SeVeDo: A Heterogeneous Transformer Accelerator for Low-Bit Inference via Hierarchical Group Quantization and SVD-Guided Mixed Precision"

0 / 0 papers shown

Title
No papers found