Distilled AI concepts, brewed by everyone.

A single, collaborative tutorial for every topic, guiding you from initial intuition to deep understanding. It functions as a collective teaching mind, where anyone can help clarify and perfect the shared knowledge.

Start reading Contribute on GitHub

The shelf

Attention Is it really all we need? The routing mechanism that helps the model to focus. i

Embeddings How a word becomes a vector, and what that vector is allowed to mean. i

Learning by Comparison Teach the model what makes two examples similar (and what makes them different) without ever looking at the labels. i

Metrics Leaving out parts of the truth: an analysis of metric estimation, theoretical assumptions, and points of structural failure in ML/DL. i

Mixture of Experts Activating only the necessary parameters, and managing the load-balancing challenges of dynamic routing. i

What next? Tell us which topic should come next, or write it yourself. Every tutorial on this shelf started as an issue.

Open like a wiki, credited like a paper.

100% open source. Refine a paragraph, fix a bug, or add a figure. Credit is baked into every contribution you make.

Contribute on GitHub Leaderboard