PagedAttention

mainstreamstable

An attention memory-management algorithm that organizes the key-value cache in blocks for efficient model serving.

雷达信号

定义

PagedAttention applies ideas from virtual memory and paging to the key-value cache used during language-model generation. It reduces fragmentation and supports cache sharing, helping a serving system use memory more effectively. The original paper introduces vLLM as a serving system built around this algorithm.

来源

Research paper first submitted to arXiv on 2023-09-12.

语境

Relevant to inference infrastructure and serving throughput. Distinguish the memory-management technique from model architecture changes or claims about answer quality. Catalog note: this is a source-verified reference entry, not a measured current-trend signal. No search-volume or growth claim is made.

来源

Efficient Memory Management for Large Language Model Serving with PagedAttention

Trend score: 0
First seen: 2023-09-12
Last seen: 2026-10-24

看今天的表

专业版打开当天的 prioritize / watch / skip 表，包括这里还不能公开的新名字。