English(EN) MoEZip: Routing-Aware KV Cache Compression for Sparse Mixture-of-Experts LLMs

MoEZip：稀疏 MoE LLM 的新型 KV 缓存压缩技术

作者 PulseAugur 编辑部 · [1 个来源] · 2026-10-23 04:00

研究人员开发了 MoEZip，一种专为稀疏专家混合（MoE）语言模型设计的新型 KV 缓存压缩技术。与可能通过改变专家选择而降低 MoE 性能的现有方法不同，MoEZip 采用了路由感知评分。该方法使用基于路由分布的 Fisher 信息矩阵的度量来估计专家选择对 KV 缓存逐出的敏感性。实验表明，在激进的压缩级别下，MoEZip 在 LongBench 和 RULER 等基准测试中的表现显著优于以前的最先进方法。 AI

影响在长上下文推理期间提高稀疏 MoE 模型的效率和性能。

排序理由详细介绍 LLM 推理优化新方法的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.LG 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: 4 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 详细介绍 LLM 推理优化新方法的学术论文。[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: paper, infra; 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: Same-day; Cluster formed today. Ranking reflects the current source set at time of score.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.LG TIER_1 English(EN) · Minsang Kim, Seungjun Baek · 2026-10-23 04:00

MoEZip: Routing-Aware KV Cache Compression for Sparse Mixture-of-Experts LLMs

arXiv:2610.30206v1 Announce Type: new Abstract: Mixture-of-Experts (MoE) language models scale model capacity efficiently by activating a small subset of parameters per token. However, in long-context inference, memory remains a major bottleneck because the KV cache grows linearl…

本文如何被排名

报道来源 [1]

MoEZip: Routing-Aware KV Cache Compression for Sparse Mixture-of-Experts LLMs

相关实体

相关话题