|国家预印本平台
首页|CAOTE: KV Caching through Attention Output Error based Token Eviction

CAOTE: KV Caching through Attention Output Error based Token Eviction

CAOTE: KV Caching through Attention Output Error based Token Eviction

来源:Arxiv_logoArxiv
英文摘要

While long context support of large language models has extended their abilities, it also incurs challenges in memory and compute which becomes crucial bottlenecks in resource-restricted devices. Token eviction, a widely adopted post-training methodology designed to alleviate the bottlenecks by evicting less important tokens from the cache, typically uses attention scores as proxy metrics for token importance. However, one major limitation of attention score as a token-wise importance metrics is that it lacks the information about contribution of tokens to the attention output. In this paper, we propose a simple eviction criterion based on the contribution of cached tokens to attention outputs. Our method, CAOTE, optimizes for eviction error due to token eviction, by seamlessly integrating attention scores and value vectors. This is the first method which uses value vector information on top of attention-based eviction scores. Additionally, CAOTE can act as a meta-heuristic method with flexible usage with any token eviction method. We show that CAOTE, when combined with the state-of-the-art attention score-based methods, always improves accuracies on the downstream task, indicating the importance of leveraging information from values during token eviction process.

Raghavv Goel、Junyoung Park、Mukul Gagrani、Dalton Jones、Matthew Morse、Harper Langston、Mingu Lee、Chris Lott

计算技术、计算机技术

Raghavv Goel,Junyoung Park,Mukul Gagrani,Dalton Jones,Matthew Morse,Harper Langston,Mingu Lee,Chris Lott.CAOTE: KV Caching through Attention Output Error based Token Eviction[EB/OL].(2025-04-18)[2025-05-15].https://arxiv.org/abs/2504.14051.点此复制

评论