English(EN) Efficient Clustering with Provable Guardrails for LLM Inference at Scale

新型聚类方法将LLM推理成本降低50倍

作者 PulseAugur 编辑部 · [1 个来源] · 2026-07-23 04:00

研究人员开发了一种新颖的两阶段聚类算法，旨在显著降低大型语言模型（LLMs）的推理成本和延迟。该方法采用Mini-batch K-Means，然后采用类似于Johnson-Chvatal启发式方法贪婪选择代表，为相似性和属性匹配提供了可证明的护栏。该算法能高效扩展到数千万个样本，其性能比标准方法高出几个数量级，并支持了一个基于角色的推荐系统投入生产，将成本降低了50倍。 AI

影响通过降低推理费用，使基于LLM的应用能够以成本效益的方式扩展到数百万用户。

排序理由该条目描述了一种新算法及其在arXiv论文中 presented 的性能基准。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv stat.ML 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: 0 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 该条目描述了一种新算法及其在arXiv论文中 presented 的性能基准。[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: 60 days old; Aged out of breaking-news scoring windows; ranking reflects the durable signal from the full source set.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv stat.ML TIER_1 English(EN) · Longshaokan Wang, Wai Tsang Keung, Punit Ghodasara, Roman Wang, Ali Dashti, Francesc Moreno-Noguer · 2026-07-23 04:00

面向大规模LLM推理的高效聚类与可证明的护栏

arXiv:2607.19704v1 Announce Type: cross Abstract: Scaling LLM-based applications to millions of users is bottlenecked by the inference cost and latency of modern foundation models. A natural fix is to cluster the inputs and call the LLM only on cluster representatives, letting ot…

本文如何被排名

报道来源 [1]

面向大规模LLM推理的高效聚类与可证明的护栏

相关实体

相关话题