1
Leyline: KV Cache Directives for Agentic Inference
突破传统KV缓存局限，专门针对Agentic LLM的推理优化，引入策略驱动编辑的新范式。
arXiv:2606.01065v1 Announce Type: cross Abstract: Modern KV cache management assumes the chatbot workload: prompts arrive once and the cache grows app…