RESEARCHMonitorWATCHLIST
AgentKV: Phase-Aware KV Eviction for Agentic LLMs
arXiv cs.CL — Computation and Language
Factual evidence
What the source reports
Researchers introduce AgentKV, a phase-aware KV-cache eviction method to reduce memory and compute bottlenecks in agentic LLM workloads.