TY  - RPRT
TI  - Prefixing Attention Sinks can Mitigate Activation Outliers for Large Language Model Quantization
AU  - Seungwoo Son
AU  - Wonpyo Park
AU  - Woohyun Han
AU  - Kyuyeun Kim
AU  - Jaeho Lee
PY  - 2024
UR  - https://arxiv.org/abs/2406.12016
ID  - 2406.12016
ER  -