<?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>架构设计 on 硅基 AGI · 智能体学习与测评</title><link>https://guijiagi.com/tags/%E6%9E%B6%E6%9E%84%E8%AE%BE%E8%AE%A1/</link><description>Recent content in 架构设计 on 硅基 AGI · 智能体学习与测评</description><generator>Hugo</generator><language>zh-cn</language><copyright>本站内容采用 CC BY-NC-SA 4.0 国际许可协议授权</copyright><lastBuildDate>Thu, 16 Jul 2026 11:19:00 +0800</lastBuildDate><atom:link href="https://guijiagi.com/tags/%E6%9E%B6%E6%9E%84%E8%AE%BE%E8%AE%A1/index.xml" rel="self" type="application/rss+xml"/><item><title>MoE架构详解：稀疏激活如何改变大模型经济学</title><link>https://guijiagi.com/posts/b1-cabd68a1/</link><pubDate>Thu, 16 Jul 2026 11:19:00 +0800</pubDate><guid>https://guijiagi.com/posts/b1-cabd68a1/</guid><description>深入解析Mixture of Experts架构的技术原理、训练方法和推理优化策略</description></item><item><title>AI Agent记忆系统架构：从短期记忆到长期持久化</title><link>https://guijiagi.com/posts/b1-f1c00d34/</link><pubDate>Thu, 16 Jul 2026 11:00:00 +0800</pubDate><guid>https://guijiagi.com/posts/b1-f1c00d34/</guid><description>深入解析AI Agent记忆系统的分层架构设计，涵盖工作记忆、情景记忆、语义记忆的工程实现</description></item><item><title>AI操作系统愿景：大模型驱动的智能计算平台</title><link>https://guijiagi.com/posts/b2-c342d404/</link><pubDate>Thu, 16 Jul 2026 10:31:00 +0800</pubDate><guid>https://guijiagi.com/posts/b2-c342d404/</guid><description>探讨以大模型为核心的AI操作系统概念，分析其架构设计、应用生态与对传统OS的颠覆潜力</description></item><item><title>多模态大模型技术演进：从CLIP到原生多模态架构</title><link>https://guijiagi.com/posts/b2-19902ccc/</link><pubDate>Thu, 16 Jul 2026 10:07:00 +0800</pubDate><guid>https://guijiagi.com/posts/b2-19902ccc/</guid><description>梳理多模态大模型从早期CLIP双塔架构到GPT-4o原生多模态的技术演进路径，分析架构设计选择与工程权衡</description></item><item><title>AI Agent记忆系统架构深度解析：从短期记忆到持久化知识库</title><link>https://guijiagi.com/posts/b2-cd38df19/</link><pubDate>Thu, 16 Jul 2026 10:00:00 +0800</pubDate><guid>https://guijiagi.com/posts/b2-cd38df19/</guid><description>深入剖析AI Agent记忆系统的三层架构设计，涵盖工作记忆、情景记忆和语义记忆的实现方案与工程实践</description></item><item><title>AI Agent的错误处理机制设计</title><link>https://guijiagi.com/posts/article-26/</link><pubDate>Sun, 12 Jul 2026 21:10:00 +0800</pubDate><guid>https://guijiagi.com/posts/article-26/</guid><description>从异常捕获到优雅降级，系统探讨AI Agent错误处理机制的设计原则与实践方案</description></item><item><title>AI Agent的错误处理机制设计</title><link>https://guijiagi.com/posts/b2-e2725e5a/</link><pubDate>Sun, 12 Jul 2026 21:10:00 +0800</pubDate><guid>https://guijiagi.com/posts/b2-e2725e5a/</guid><description>从异常捕获到优雅降级，系统探讨AI Agent错误处理机制的设计原则与实践方案</description></item><item><title>从零搭建企业级RAG系统：完整方案设计</title><link>https://guijiagi.com/posts/article-12/</link><pubDate>Sun, 12 Jul 2026 18:50:00 +0800</pubDate><guid>https://guijiagi.com/posts/article-12/</guid><description>手把手指导企业级RAG系统的架构设计、技术选型和部署方案，覆盖从数据接入到线上监控的全流程</description></item><item><title>从零搭建企业级RAG系统：完整方案设计</title><link>https://guijiagi.com/posts/b2-f0acee82/</link><pubDate>Sun, 12 Jul 2026 18:50:00 +0800</pubDate><guid>https://guijiagi.com/posts/b2-f0acee82/</guid><description>手把手指导企业级RAG系统的架构设计、技术选型和部署方案，覆盖从数据接入到线上监控的全流程</description></item><item><title>深度解析Transformer-XL：长序列建模的突破</title><link>https://guijiagi.com/posts/article-06/</link><pubDate>Sun, 12 Jul 2026 17:50:00 +0800</pubDate><guid>https://guijiagi.com/posts/article-06/</guid><description>从原理到实践，全面解析Transformer-XL如何突破标准Transformer的序列长度限制</description></item><item><title>深度解析Transformer-XL：长序列建模的突破</title><link>https://guijiagi.com/posts/b2-5c0c8534/</link><pubDate>Sun, 12 Jul 2026 17:50:00 +0800</pubDate><guid>https://guijiagi.com/posts/b2-5c0c8534/</guid><description>从原理到实践，全面解析Transformer-XL如何突破标准Transformer的序列长度限制</description></item><item><title>Agent记忆系统设计：从短期到长期的演进</title><link>https://guijiagi.com/posts/article-04/</link><pubDate>Sun, 12 Jul 2026 17:30:00 +0800</pubDate><guid>https://guijiagi.com/posts/article-04/</guid><description>系统性地探讨AI Agent记忆系统的架构设计，覆盖工作记忆、情景记忆和长期记忆的实现方案</description></item><item><title>Agent记忆系统设计：从短期到长期的演进</title><link>https://guijiagi.com/posts/b2-c06ea946/</link><pubDate>Sun, 12 Jul 2026 17:30:00 +0800</pubDate><guid>https://guijiagi.com/posts/b2-c06ea946/</guid><description>系统性地探讨AI Agent记忆系统的架构设计，覆盖工作记忆、情景记忆和长期记忆的实现方案</description></item><item><title>Agentic RAG：当RAG遇到智能体的架构革命</title><link>https://guijiagi.com/posts/agentic-rag-architecture-revolution/</link><pubDate>Tue, 30 Jun 2026 09:40:00 +0800</pubDate><guid>https://guijiagi.com/posts/agentic-rag-architecture-revolution/</guid><description>Agentic RAG将检索增强生成从被动管道升级为主动决策系统，本文深入解析其架构设计、核心模式与工程实践</description></item><item><title>多模态RAG：图文混合检索的架构设计</title><link>https://guijiagi.com/posts/multimodal-rag-architecture/</link><pubDate>Tue, 30 Jun 2026 09:40:00 +0800</pubDate><guid>https://guijiagi.com/posts/multimodal-rag-architecture/</guid><description>多模态RAG让AI能同时理解和检索图片与文字，本文详解四种架构设计与Claude 4/GPT-5的实践方案</description></item><item><title>Agent编排模式2026：从串行到图式的完整设计指南</title><link>https://guijiagi.com/posts/agent-orchestration-patterns-2026/</link><pubDate>Tue, 30 Jun 2026 09:30:00 +0800</pubDate><guid>https://guijiagi.com/posts/agent-orchestration-patterns-2026/</guid><description>系统梳理2026年Agent编排的核心模式，从简单串行到复杂图式编排，包含代码示例、选型决策树和反模式警示</description></item><item><title>Agent记忆系统设计：短期、长期与情景记忆的实现</title><link>https://guijiagi.com/posts/agent-memory-system-design/</link><pubDate>Tue, 30 Jun 2026 09:30:00 +0800</pubDate><guid>https://guijiagi.com/posts/agent-memory-system-design/</guid><description>系统讲解Agent记忆系统的三层架构设计，涵盖短期工作记忆、长期语义记忆和情景记忆的实现方案与工程实践</description></item><item><title>Transformer架构2026：从注意力机制到混合专家的演进</title><link>https://guijiagi.com/posts/transformer-architecture-2026-attention-to-moe-evolution/</link><pubDate>Tue, 30 Jun 2026 09:20:00 +0800</pubDate><guid>https://guijiagi.com/posts/transformer-architecture-2026-attention-to-moe-evolution/</guid><description>深入解析Transformer架构在2026年的最新演进，涵盖注意力机制优化、MoE混合专家、线性注意力等前沿方向</description></item><item><title>Transformer架构2026：从注意力机制到混合专家的演进</title><link>https://guijiagi.com/posts/transformer-architecture-2026/</link><pubDate>Tue, 30 Jun 2026 09:20:00 +0800</pubDate><guid>https://guijiagi.com/posts/transformer-architecture-2026/</guid><description>深入解析Transformer架构在2026年的最新演进，涵盖注意力机制优化、混合专家模型、线性注意力等前沿方向</description></item><item><title>Transformer 架构 2026 最新演进：从 Attention 到 MoE 再到 Mamba</title><link>https://guijiagi.com/posts/transformer-architecture-2026-evolution/</link><pubDate>Sun, 28 Jun 2026 11:00:00 +0800</pubDate><guid>https://guijiagi.com/posts/transformer-architecture-2026-evolution/</guid><description>深入解析Transformer架构在2026年的最新演进，涵盖Attention机制优化、MoE混合专家架构、Mamba状态空间模型三大方向</description></item><item><title>Agentic RAG：当 RAG 遇到 Agent，检索增强的下一步</title><link>https://guijiagi.com/posts/agentic-rag-when-rag-meets-agent/</link><pubDate>Sun, 28 Jun 2026 10:00:00 +0800</pubDate><guid>https://guijiagi.com/posts/agentic-rag-when-rag-meets-agent/</guid><description>探索 Agentic RAG 的设计理念、架构模式和工程实现，理解为什么 Agent 是 RAG 的自然进化方向</description></item><item><title>RAG 架构 2026 最新实践：从 Naive RAG 到 GraphRAG+Agent</title><link>https://guijiagi.com/posts/rag-architecture-2026-naive-to-graphrag-agent/</link><pubDate>Sun, 28 Jun 2026 10:00:00 +0800</pubDate><guid>https://guijiagi.com/posts/rag-architecture-2026-naive-to-graphrag-agent/</guid><description>系统梳理 RAG 架构演进，从最基础的 Naive RAG 到 2026 年最前沿的 GraphRAG+Agent 范式，附完整架构对比与代码示例</description></item><item><title>RAG 架构设计模式：从朴素 RAG 到模块化 RAG 的演进</title><link>https://guijiagi.com/posts/rag-architecture-patterns/</link><pubDate>Wed, 24 Jun 2026 10:10:00 +0800</pubDate><guid>https://guijiagi.com/posts/rag-architecture-patterns/</guid><description>系统梳理 RAG 架构的演进路径、设计模式和工程选型</description></item><item><title>Agent 记忆系统 2026：从短期上下文到持久记忆的工程实践</title><link>https://guijiagi.com/posts/agent-%E8%AE%B0%E5%BF%86%E7%B3%BB%E7%BB%9F-2026-%E4%BB%8E%E7%9F%AD%E6%9C%9F%E4%B8%8A%E4%B8%8B%E6%96%87%E5%88%B0%E6%8C%81%E4%B9%85%E8%AE%B0%E5%BF%86%E7%9A%84%E5%B7%A5%E7%A8%8B%E5%AE%9E%E8%B7%B5/</link><pubDate>Sat, 20 Jun 2026 00:00:00 +0800</pubDate><guid>https://guijiagi.com/posts/agent-%E8%AE%B0%E5%BF%86%E7%B3%BB%E7%BB%9F-2026-%E4%BB%8E%E7%9F%AD%E6%9C%9F%E4%B8%8A%E4%B8%8B%E6%96%87%E5%88%B0%E6%8C%81%E4%B9%85%E8%AE%B0%E5%BF%86%E7%9A%84%E5%B7%A5%E7%A8%8B%E5%AE%9E%E8%B7%B5/</guid><description>深度解析 2026 年 Agent 记忆系统的最佳实践，涵盖向量数据库、知识图谱、分层记忆架构与端到端实现方案</description></item></channel></rss>