TY  - RPRT
TI  - Enhancing Ultra-Low-Bit Quantization of Large Language Models Through Saliency-Aware Partial Retraining
AU  - Deyu Cao
AU  - Samin Aref
PY  - 2025
DO  - 10.1007/978-3-032-00891-6_28
UR  - https://arxiv.org/abs/2504.13932
ID  - 2504.13932
ER  -