@article{sparqattentionbandwidthefficientllm,
  title = {SparQ Attention: Bandwidth-Efficient LLM Inference},
  author = {Luka Ribar and Ivan Chelombiev and Luke Hudlass-Galley and Charlie Blake and Carlo Luschi and Douglas Orr},
  year = {2023},
  eprint = {2312.04985},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2312.04985v6},
}