10 Must Read ML Blog Posts
A collection of high-impact machine learning blog posts.
I have been doing NLP/ML research for the last 6 years. I have come across a lot of machine learning resources and papers. Today, I kept thinking about the machine learning / NLP / deep learning related blog posts (not papers) that have been transformational for me. In this blog post, I provide a short collection of a few high-impact blog posts that come to mind.
This post was originally a Twitter thread.
1) The Unreasonable Effectiveness of Recurrent Neural Networks
by Andrej Karpathy
🔗 link
2) Understanding Convolutional Neural Networks for NLP
by Denny Britz
🔗 link
3) Language Models
by Vered Shwartz
🔗 link
4) An overview of gradient descent optimization algorithms
by Sebastian Ruder
🔗 link
5) Understanding LSTM Networks
by Chris Olah
🔗 link
6) Deep Learning: Our Miraculous Year 1990-1991
by Jürgen Schmidhuber
🔗 link
7) An Overview of Deep Learning for Curious People
by Lilian Weng
🔗 link
8) Attention and Memory in Deep Learning and NLP
by Denny Britz
🔗 link
9) Neural Networks and Deep Learning
by Michael Nielsen
🔗 link
10) Troubling Trends in Machine Learning Scholarship
by Zachary Lipton
🔗 link
Bonus….
11) The Illustrated Transformer
by Jay Alammar
🔗 link
The list will be different for every generation but these are some of the blog posts that had been transformational for me. There are many more high-quality blog posts out there, so feel free to share them in the comments. I thank these and many other people for making ML more accessible and welcoming.
As mentioned in the comments, it’s important to note that these posts might be relatively old and outdated.
To get regular updates on new ML and NLP resources, follow me on Twitter.
I cannot help pointing out that some of the resources are old, and hence, outdated (not all old posts are outdated).
I went to Lillian Weng's blog post, which was written in 2017, and which mentions Tensorflow as a new library, and there is no mention of PyTorch at all. Now PyTorch has overtaken TensorFlow for research, competition, and educational purposes, and the future of JAX looks bright.
This is just one example. I think it would be better if you point out the limitations of some of the resources here for their being old and outdated.
This is a great blog post - So clear and easy to follow. All your hard work is so much appreciated.