TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Understanding LSTM networks

99 pointsby michael_nielsenover 9 years ago

6 comments

reader5000over 9 years ago
Whats amazing to me is that if I understand correctly backprop still works. It is very odd that SGD on the error function for some training data is conceptually equivalent to teaching all the gates for each hidden feature when to open/close given the next input in a sequence.
评论 #10131910 未加载
otover 9 years ago
colah, your posts combine a deep level of understanding and an exceptional clarity. These are both rare, especially in the cargo-cult-driven world of neural networks and deep learning.<p>I hope you keep writing as much as you can. Thanks!
d136oover 9 years ago
Thanks colah, that was a very readable walk-through. I&#x27;ve been making my way through Bishop&#x27;s PRML ch 5 to get as much of a handle as possible on NNs, but your intro here to LSTM&#x27;s makes me want to jump ahead and skip to the new stuff :)
p1eskover 9 years ago
Michael, this post nicely completes your book about neural networks. I was a little surprised you didn&#x27;t write it yourself.
ambicapterover 9 years ago
Anybody know what he uses for his diagrams?
评论 #10131887 未加载
mistermasterover 9 years ago
great explanation. many thanks! hochreiter is genius!
评论 #10131362 未加载