<?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>PyTorch on Blog</title><link>https://icelemon.top/tags/pytorch/</link><description>Recent content in PyTorch on Blog</description><generator>Hugo</generator><language>zh-CN</language><lastBuildDate>Sat, 24 Oct 2026 00:00:00 +0800</lastBuildDate><atom:link href="https://icelemon.top/tags/pytorch/index.xml" rel="self" type="application/rss+xml"/><item><title>机器人具身学习路线</title><link>https://icelemon.top/tech/robotics/embodied-learning-route/</link><pubDate>Sat, 24 Oct 2026 00:00:00 +0800</pubDate><guid>https://icelemon.top/tech/robotics/embodied-learning-route/</guid><description>先建立机器人与学习基础，再完成BC、策略梯度和闭环评测，理解视觉动作模型的接口与研究边界。</description></item><item><title>模仿学习与行为克隆</title><link>https://icelemon.top/tech/robotics/imitation-learning/</link><pubDate>Sat, 24 Oct 2026 00:00:00 +0800</pubDate><guid>https://icelemon.top/tech/robotics/imitation-learning/</guid><description>把演示转成观测—动作数据，理解行为克隆的损失、闭环分布偏移和按episode切分。</description></item><item><title>强化学习：MDP、策略梯度与机器人约束</title><link>https://icelemon.top/tech/robotics/reinforcement-learning/</link><pubDate>Sat, 24 Oct 2026 00:00:00 +0800</pubDate><guid>https://icelemon.top/tech/robotics/reinforcement-learning/</guid><description>明确状态、动作、奖励和终止，用可运行REINFORCE实验理解回报、基线与闭环评估。</description></item></channel></rss>