@article{compresskvsemanticretrievalguidedkvcache,
  title = {CompressKV: Semantic-Retrieval-Guided KV-Cache Compression for Resource-Efficient Long-Context LLM Inference},
  author = {Xiaolin Lin and Jingcun Wang and Olga Kondrateva and Yiyu Shi and Bing Li and Grace Li Zhang},
  year = {2026},
  eprint = {2606.24467},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2606.24467},
}