RotateKV: Accurate and Robust 2-Bit KV Cache Quantization for LLMs via Outlier-Aware Adaptive Rotations

Su et al. "RotateKV: Accurate and Robust 2-Bit KV Cache Quantization for LLMs via Outlier-Aware Adaptive Rotations." International Joint Conference on Artificial Intelligence, 2025. doi:10.24963/IJCAI.2025/690

[Su et al. "RotateKV: Accurate and Robust 2-Bit KV Cache Quantization for LLMs via Outlier-Aware Adaptive Rotations." International Joint Conference on Artificial Intelligence, 2025.](https://mlanthology.org/ijcai/2025/su2025ijcai-rotatekv/) doi:10.24963/IJCAI.2025/690

@inproceedings{su2025ijcai-rotatekv,
  title     = {{RotateKV: Accurate and Robust 2-Bit KV Cache Quantization for LLMs via Outlier-Aware Adaptive Rotations}},
  author    = {Su, Zunhai and Wei, Hanyu and Chen, Zhe and Shen, Wang and Li, Linge and Yu, Huangqi and Yuan, Kehong},
  booktitle = {International Joint Conference on Artificial Intelligence},
  year      = {2025},
  pages     = {6200-6208},
  doi       = {10.24963/IJCAI.2025/690},
  url       = {https://mlanthology.org/ijcai/2025/su2025ijcai-rotatekv/}
}