KV cache

2 posts tagged with "KV cache"

July 23, 2026

InfoFlow KV: Information-Flow-Aware KV Recomputation for Long Context

A selective KV recomputation method using attention-norm criterion and inference-consistent RoPE ordering for long-context LLM/VLM inference

KV cache long context RAG RoPE inference efficiency selective recomputation

July 13, 2026

AAFLOW+: Stateful Operator Abstraction with Zero-Copy Distributed KV Cache Orchestration for Multi-Agent Workflows

A stateful extension of agentic workflow operators that elevates KV cache to a first-class distributed systems object, enabling zero-copy, transfer-aware execution for multi-agent LLM workflows.

multi-agent KV cache distributed systems LLM serving stateful operator AAFLOW zero-copy agentic workflow