<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>公众号文章 on LiuYang Lab</title><link>https://jaywillow0.github.io/wechat/</link><description>Recent content in 公众号文章 on LiuYang Lab</description><generator>Hugo</generator><language>zh-CN</language><copyright>© 2026 LiuYang</copyright><lastBuildDate>Sun, 30 Aug 2026 16:01:47 +0800</lastBuildDate><atom:link href="https://jaywillow0.github.io/wechat/index.xml" rel="self" type="application/rss+xml"/><item><title>创建个人风格的Skill 初尝试</title><link>https://jaywillow0.github.io/wechat/personal-skill-first-attempt/</link><pubDate>Sun, 30 Aug 2026 16:01:47 +0800</pubDate><guid>https://jaywillow0.github.io/wechat/personal-skill-first-attempt/</guid><description>&lt;p&gt;一次真实建模任务的完整复盘，记录如何审查过于乐观的模型结果，并把踩过的坑整理成可复用的工程规则。&lt;/p&gt;</description></item><item><title>从昂贵仿真到Pareto前沿，一套可复用的LFP电池代理优化方法</title><link>https://jaywillow0.github.io/wechat/lfp-surrogate-pareto-optimization/</link><pubDate>Fri, 07 Aug 2026 23:17:08 +0800</pubDate><guid>https://jaywillow0.github.io/wechat/lfp-surrogate-pareto-optimization/</guid><description>&lt;p&gt;围绕比能量与比功率的权衡，梳理高保真模型、代理模型、多目标寻优和物理回代验证组成的工程闭环。&lt;/p&gt;</description></item><item><title>Vibe Coding实战：制造业仿真工程师如何从零搭建 AutoML 全栈平台</title><link>https://jaywillow0.github.io/wechat/vibe-coding-automl-platform/</link><pubDate>Sat, 25 Jul 2026 09:33:45 +0800</pubDate><guid>https://jaywillow0.github.io/wechat/vibe-coding-automl-platform/</guid><description>&lt;p&gt;文章重点复盘需求拆分、工程边界、验证方式与人机协作，而不是罗列一个会过时的代码清单。&lt;/p&gt;</description></item><item><title>前100圈预测锂电池寿命：Severson 2019 经典工作的复现与工业迁移边界的思考</title><link>https://jaywillow0.github.io/wechat/severson-2019-battery-life-reproduction/</link><pubDate>Mon, 13 Jul 2026 19:01:18 +0800</pubDate><guid>https://jaywillow0.github.io/wechat/severson-2019-battery-life-reproduction/</guid><description>&lt;p&gt;从论文复现走向工程迁移，重点辨析快充研究数据与低倍率、长寿命工业数据之间的域差异。&lt;/p&gt;</description></item></channel></rss>