How should we think about neural networks? How can they help us think, ourselves?
* An exploration of very general techniques for visualizing neural networks.
* Visualizations of sophisticated models from Google's deep learning research group.
* meta-SNE, a new algorithm for visualizing entire spaces of neural networks, instead of individual ones.
* A visualization of English Wikipedia, as a whole!
* A way to use deep learning for gender studies (and more)!