MixKVQ: Query-Aware Mixed-Precision KV Cache Quantization for Long-Context Reasoning

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Zhang, Tao, Zeng, Ziqian, Peng, Hao, Zhuang, Huiping, Chen, Cen
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866918258992480256
author Zhang, Tao
Zeng, Ziqian
Peng, Hao
Zhuang, Huiping
Chen, Cen
author_facet Zhang, Tao
Zeng, Ziqian
Peng, Hao
Zhuang, Huiping
Chen, Cen
contents Long Chain-of-Thought (CoT) reasoning has significantly advanced the capabilities of Large Language Models (LLMs), but this progress is accompanied by substantial memory and latency overhead from the extensive Key-Value (KV) cache. Although KV cache quantization is a promising compression technique, existing low-bit quantization methods often exhibit severe performance degradation on complex reasoning tasks. Fixed-precision quantization struggles to handle outlier channels in the key cache, while current mixed-precision strategies fail to accurately identify components requiring high-precision representation. We find that an effective low-bit KV cache quantization strategy must consider two factors: a key channel's intrinsic quantization difficulty and its relevance to the query. Based on this insight, we propose MixKVQ, a novel plug-and-play method that introduces a lightweight, query-aware algorithm to identify and preserve critical key channels that need higher precision, while applying per-token quantization for value cache. Experiments on complex reasoning datasets demonstrate that our approach significantly outperforms existing low-bit methods, achieving performance comparable to a full-precision baseline at a substantially reduced memory footprint.
format Preprint
id arxiv_https___arxiv_org_abs_2512_19206
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle MixKVQ: Query-Aware Mixed-Precision KV Cache Quantization for Long-Context Reasoning
Zhang, Tao
Zeng, Ziqian
Peng, Hao
Zhuang, Huiping
Chen, Cen
Machine Learning
Artificial Intelligence
Long Chain-of-Thought (CoT) reasoning has significantly advanced the capabilities of Large Language Models (LLMs), but this progress is accompanied by substantial memory and latency overhead from the extensive Key-Value (KV) cache. Although KV cache quantization is a promising compression technique, existing low-bit quantization methods often exhibit severe performance degradation on complex reasoning tasks. Fixed-precision quantization struggles to handle outlier channels in the key cache, while current mixed-precision strategies fail to accurately identify components requiring high-precision representation. We find that an effective low-bit KV cache quantization strategy must consider two factors: a key channel's intrinsic quantization difficulty and its relevance to the query. Based on this insight, we propose MixKVQ, a novel plug-and-play method that introduces a lightweight, query-aware algorithm to identify and preserve critical key channels that need higher precision, while applying per-token quantization for value cache. Experiments on complex reasoning datasets demonstrate that our approach significantly outperforms existing low-bit methods, achieving performance comparable to a full-precision baseline at a substantially reduced memory footprint.
title MixKVQ: Query-Aware Mixed-Precision KV Cache Quantization for Long-Context Reasoning
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2512.19206