Skip to content
conftrace_
2025 ICML ICML 2025

CateKV: On Sequential Consistency for Long-Context LLM Inference Acceleration

Abstract

No abstract available. No abstract was captured for this paper when its proceedings were scraped - most often an older edition published without one online. How the corpus is built.

Authors

Haoyun Jiang, Haolin Li, Jianwei Zhang, Fei Huang, Qiang Hu, Minmin Sun, Shuai Xiao, Yong Li, Junyang Lin, Jiangchao Yao

Topics

Machine Learning > Application Areas > Efficient Computing Artificial Intelligence > Core AI > Large Language Models

Keywords

memory usage attention head inference latency coefficient-of-variation-based algorithm hybrid kv cache sequential consistency
Download PDF