<?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>Agent评估 on 硅基 AGI · 智能体学习与测评</title><link>https://guijiagi.com/tags/agent%E8%AF%84%E4%BC%B0/</link><description>Recent content in Agent评估 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>Mon, 13 Jul 2026 05:10:00 +0800</lastBuildDate><atom:link href="https://guijiagi.com/tags/agent%E8%AF%84%E4%BC%B0/index.xml" rel="self" type="application/rss+xml"/><item><title>从零开始构建AI Agent评估管线</title><link>https://guijiagi.com/posts/article-74/</link><pubDate>Mon, 13 Jul 2026 05:10:00 +0800</pubDate><guid>https://guijiagi.com/posts/article-74/</guid><description>系统性介绍AI Agent评估管线的构建方法，从指标设计到数据集构建到自动化评估全流程实操指南</description></item><item><title>AI Agent的可靠性评估：从单点到系统级</title><link>https://guijiagi.com/posts/article-07/</link><pubDate>Sun, 12 Jul 2026 18:00:00 +0800</pubDate><guid>https://guijiagi.com/posts/article-07/</guid><description>构建系统化的Agent可靠性评估框架，从任务完成率到错误恢复能力的全方位测评方法论</description></item><item><title>AI Agent的可靠性评估：从单点到系统级</title><link>https://guijiagi.com/posts/b2-68db480e/</link><pubDate>Sun, 12 Jul 2026 18:00:00 +0800</pubDate><guid>https://guijiagi.com/posts/b2-68db480e/</guid><description>构建系统化的Agent可靠性评估框架，从任务完成率到错误恢复能力的全方位测评方法论</description></item><item><title>Agent基准设计2026：如何评估智能体能力</title><link>https://guijiagi.com/posts/agent-benchmark-design-2026/</link><pubDate>Thu, 02 Jul 2026 11:16:00 +0800</pubDate><guid>https://guijiagi.com/posts/agent-benchmark-design-2026/</guid><description>2026年Agent能力评估基准设计方法，覆盖工具调用、规划与多步推理</description></item><item><title>Agent 基准测试方法论：从设计到执行</title><link>https://guijiagi.com/posts/agent-benchmark-methodology/</link><pubDate>Wed, 24 Jun 2026 14:00:00 +0800</pubDate><guid>https://guijiagi.com/posts/agent-benchmark-methodology/</guid><description>系统讲解 Agent 基准测试的设计原则、测试集构建、评估指标体系及主流基准对比</description></item></channel></rss>