(C-x means ctrl+x, M-x means alt+x)

The default prefix is C-b. If you (or your muscle memory) prefer C-a, you need to add this to ~/.tmux.conf:

• This document was initally written for a friend of mine, Jiawen Li, so it might reflect some personal tastes here and there. For instance, some discussions are geared towards Windows, though *nix is obviously superior. For another example, when I say "you seem to love Sublime Text a lot," I'm certainly not expecting most people to love Sublime (in fact I never used it for more than three minutes in a row).

• This document is written in Markdown. The Markdown rendering engine on GitHub Gist is somewhat limited and easily confused, so there might be some rendering errors not intended by the author. For instance, in the section "Useful stuff for LaTeX beginners (and experts, occasionally)", the outermost numbered list should be numbered 1, 2, 3, ... But it is rendered instead as 1, 1, 2, ... on GitHub Gist due to the complicated stuff in between. (Maybe it's my syntax that's sloppy, but the document is rendered correctly on all of my local engin

A comparison of Theano with other deep learning frameworks, highlighting a series of low-level design choices in no particular order.

Symbolic: Theano, CGT; Automatic: Torch, MXNet

Symbolic and automatic differentiation are often confused or used interchangeably, although their implementations are significantly different.

Ideas are cheap. Make a prototype, sketch a CLI session, draw a wireframe. Discuss around concrete examples, not hand-waving abstractions. Don't say you did something, provide a URL that proves it.

Nothing is real until it's being used by a real user. This doesn't mean you make a prototype in the morning and blog about it in the evening. It means you find one person you believe your product will help and try to get them to use it.

1. Feature Learning

• Learning Feature Representations with K-means by Adam Coates and Andrew Y. Ng
• The devil is in the details: an evaluation of recent feature encoding methods by Chatfield et. al.
• Emergence of Object-Selective Features in Unsupervised Feature Learning by Coates, Ng
• Scaling Learning Algorithms towards AI Benjio & LeCun

2. Deep Learning

• Dropout: A Simple Way to Prevent Neural Networks from Overfitting by Nitish Srivastava, Geoffrey Hinton, Alex Krizhevsky, Ilya Sutskever and Ruslan Salakhutdinov
• [Understanding the difficulty of training deep feedforward neural networks](http://jmlr.org/proceedings/papers/v9/glorot10a/gloro