<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>MoE on MessageDaily</title><link>https://inkeast.github.io/MessageDaily/tags/moe/</link><description>Recent content in MoE on MessageDaily</description><generator>Hugo</generator><language>zh-cn</language><lastBuildDate>Wed, 26 Aug 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://inkeast.github.io/MessageDaily/tags/moe/index.xml" rel="self" type="application/rss+xml"/><item><title>领读Kimi K3技术报告：一个清华架构博士眼中的注意力谱系与「有效scaling」</title><link>https://inkeast.github.io/MessageDaily/posts/2026-08-26-kimi-k3-tech-report-architecture-lead-read/</link><pubDate>Wed, 26 Aug 2026 00:00:00 +0000</pubDate><guid>https://inkeast.github.io/MessageDaily/posts/2026-08-26-kimi-k3-tech-report-architecture-lead-read/</guid><description>一集面向技术读者的Kimi K3技术报告领读播客，嘉宾孙宇涛（清华计算机系博士生、上海创智学院pre-doc，研究方向LLM架构与预训练）从K3出发串联起十多篇前作，把KDA线性注意力的每一项公式还原成RetNet→Mamba→DeltaNet→Gated DeltaNet的历史叠加，讲清channel-wise衰减、low-rank dk与BF16 tile的kernel co-design，MLA+QK-norm式门控的稳定性逻辑，Latent MoE对通信开销的削减，以及Quantile Balancing如何用线性规划一步求出负载均衡bias。预训练侧K3反潮流回归cosine decay、在混合注意力里用NoPE让长上下文免调参外推；后训练侧on-policy蒸馏成为多teacher多reward的「多模型合板」方案。嘉宾的暴论：大模型架构没有本质创新了，K3最核心的变量是size——2.8T总参、百B激活、K2的2.5倍scaling效率，而把size做work才是真创新。</description></item></channel></rss>