English(EN) Quality-Constrained Routing over a Fixed Pool of Quantized Mixture-of-Experts Instances

新的FWP路由策略优化量化MoE模型

作者 PulseAugur 编辑部 · [1 个来源] · 2026-09-14 04:00

研究人员开发了一种新颖的量化专家混合（MoE）模型路由策略，旨在优化吞吐量同时管理质量下降。这种新方法称为 Fragility-Weighted Perplexity (FWP)，通过分析提示（prompt）令牌并将其校准到候选模型实例来预测特定请求的风险。这种方法允许更有效地利用预先实例化的固定MoE实例池，在离线评估中优于静态或与请求无关的混合策略。 AI

影响引入了一种更有效的MoE模型路由机制，有望提高推理速度和成本效益。

排序理由学术论文，详细介绍了一种优化MoE模型路由的新方法。[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: 24 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 学术论文，详细介绍了一种优化MoE模型路由的新方法。[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: 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: 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.LG TIER_1 English(EN) · Zhenghong Huang, Hongfan Wu, Jiheng Zhang · 2026-09-14 04:00

面向固定量化混合专家实例池的质量约束路由

arXiv:2609.12550v1 Announce Type: new Abstract: Quantized Mixture-of-Experts (MoE) services can hold several pre-materialized instances of one base model, but quantization damage varies sharply across requests and bitwidths. Because instance materialization and replica counts con…

本文如何被排名

报道来源 [1]

面向固定量化混合专家实例池的质量约束路由

相关实体

相关话题