@article{intactkvimprovinglargelanguagemodel, title = {IntactKV: Improving Large Language Model Quantization by Keeping Pivot Tokens Intact}, author = {Ruikang Liu and Haoli Bai and Haokun Lin and Yuening Li and Han Gao and Zhengzhuo Xu and Lu Hou and Jun Yao and Chun Yuan}, year = {2024}, eprint = {2403.01241}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2403.01241v2}, }