@article{locretenhancingevictioninlongcontextllm, title = {Locret: Enhancing Eviction in Long-Context LLM Inference with Trained Retaining Heads on Consumer-Grade Devices}, author = {Yuxiang Huang and Binhang Yuan and Xu Han and Chaojun Xiao and Zhiyuan Liu}, year = {2024}, eprint = {2410.01805}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2410.01805v2}, }