<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Seq2seq on 111qqz的小窝</title><link>https://111qqz.com/tags/seq2seq/</link><description>Recent content in Seq2seq on 111qqz的小窝</description><generator>Hugo -- gohugo.io</generator><language>zh-cn</language><managingEditor>hust.111qqz@gmail.com (111qqz)</managingEditor><webMaster>hust.111qqz@gmail.com (111qqz)</webMaster><copyright>© 2011-2026 111qqz</copyright><lastBuildDate>Sun, 13 Sep 2026 15:00:00 +0800</lastBuildDate><atom:link href="https://111qqz.com/tags/seq2seq/index.xml" rel="self" type="application/rss+xml"/><item><title>从 RNN / LSTM / GRU 到早期 Attention：为什么“压缩历史”最终变成了“按需读取”</title><link>https://111qqz.com/2026/09/seq2seq-to-attention/</link><pubDate>Sun, 13 Sep 2026 15:00:00 +0800</pubDate><author>hust.111qqz@gmail.com (111qqz)</author><guid>https://111qqz.com/2026/09/seq2seq-to-attention/</guid><description>&lt;p&gt;上一篇&lt;a href="https://111qqz.com/2026/09/rnn-lstm-gru-architecture/" &gt;《从 RNN 到 LSTM，再到 GRU》&lt;/a&gt;把 gate 机制理了一遍：LSTM / GRU 把 recurrent state 的信息流变成了可学习的动态控制，决定什么写入、什么保留、什么暴露。写到那里，我以为 RNN 这条线差不多收尾了。&lt;/p&gt;</description></item></channel></rss>