@article{moskamixtureofsharedkvattentionforeffici,
  title = {MoSKA: Mixture of Shared KV Attention for Efficient Long-Sequence LLM Inference},
  author = {Myunghyun Rhee and Sookyung Choi and Euiseok Kim and Joonseop Sim and Youngpyo Joo and Hoshik Kim},
  year = {2025},
  eprint = {2511.06010},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2511.06010},
}