<?xml version='1.0' encoding='utf-8'?>
<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9" xmlns:video="http://www.google.com/schemas/sitemap-video/1.1"><url><loc>https://bdskb.com/ai/watch/neural-networks/</loc><video:video><video:thumbnail_loc>https://bdskb.com/ai/assets/covers/neural-networks.png</video:thumbnail_loc><video:title>神经网络是什么？从一张手写数字看懂 AI</video:title><video:description>从手写数字识别入手，先理解输入、权重和输出，再看复杂模型。</video:description><video:player_loc>https://player.bilibili.com/player.html?bvid=BV1bx411M7Zx&amp;p=1&amp;autoplay=0&amp;danmaku=0</video:player_loc></video:video></url><url><loc>https://bdskb.com/ai/watch/gradient-descent/</loc><video:video><video:thumbnail_loc>https://bdskb.com/ai/assets/covers/gradient-descent.png</video:thumbnail_loc><video:title>梯度下降：神经网络到底怎样学习</video:title><video:description>用“沿山坡下行”的直觉，理解训练为什么要计算误差和调整参数。</video:description><video:player_loc>https://player.bilibili.com/player.html?bvid=BV1Ux411j7ri&amp;p=1&amp;autoplay=0&amp;danmaku=0</video:player_loc></video:video></url><url><loc>https://bdskb.com/ai/watch/transformers/</loc><video:video><video:thumbnail_loc>https://bdskb.com/ai/assets/covers/transformers.png</video:thumbnail_loc><video:title>Transformer 是什么？看懂 GPT 生成文字的流程</video:title><video:description>先建立整条计算流程，再区分词元、向量、注意力与输出概率。</video:description><video:player_loc>https://player.bilibili.com/player.html?bvid=BV13z421U7cs&amp;p=1&amp;autoplay=0&amp;danmaku=0</video:player_loc></video:video></url><url><loc>https://bdskb.com/ai/watch/attention/</loc><video:video><video:thumbnail_loc>https://bdskb.com/ai/assets/covers/attention.png</video:thumbnail_loc><video:title>注意力机制：AI 怎样根据上下文理解一个词</video:title><video:description>看清 Q、K、V 各做什么，理解词元如何从其他位置读取信息。</video:description><video:player_loc>https://player.bilibili.com/player.html?bvid=BV1TZ421j7Ke&amp;p=1&amp;autoplay=0&amp;danmaku=0</video:player_loc></video:video></url></urlset>