English(EN) OSCAR: Offline Spectral Covariance-Aware Rotation for 2-bit KV Cache Quantization

新的OSCAR方法将LLM KV缓存大小削减8倍，提升速度

作者 PulseAugur 编辑部 · [1 个来源] · 2026-11-14 05:00

研究人员开发了OSCAR，一种将KV缓存量化到2位精度的新颖方法，显著减少了大型语言模型的内存使用并提高了推理速度。该技术将量化基与注意力机制对齐，即使在极度压缩的缓存下也能保持准确性。OSCAR已集成到SGLang和vLLM等服务引擎中，在各种基准测试和模型规模上均显示出吞吐量和解码速度的显著提升。 AI

影响通过减少内存占用和提高推理速度，实现了更高效的大型语言模型部署。

排序理由该条目描述了在arXiv论文中提出的一种新颖的KV缓存量化方法。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.AI 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 15 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 该条目描述了在arXiv论文中提出的一种新颖的KV缓存量化方法。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: infra, model release; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Breaking (< 6h); Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.AI TIER_1 English(EN) · Zhongzhu Zhou, Donglin Zhuang, Jisen Li, Ziyan Chen, Shuaiwen Leon Song, Ben Athiwaratkun, Xiaoxia Wu · 2026-11-14 05:00

OSCAR：用于 2 位 KV 缓存量化的离线光谱协方差感知旋转

arXiv:2605.17757v2 Announce Type: replace-cross Abstract: INT2 KV-cache quantization is attractive for long-context LLM serving, yet making it both accurate and practical remains challenging. Data-oblivious rotations such as Hadamard transforms flatten channel outliers, but still…

本文如何被排名

报道来源 [1]

OSCAR：用于 2 位 KV 缓存量化的离线光谱协方差感知旋转

相关实体

相关话题