About

The Posterior is a blog about learning machine learning, AI and statistics. The focus is on getting to the core of an idea: why it works, where it comes from, and where it breaks.

What you’ll find here

• Explainers on ML, AI and statistics, with code and figures where they help.
• Paper reviews and overviews of work worth reading.
• Opinions on tech and AI, now and then. Everyone has some; we try to keep ours rare and backed by evidence.
• Advice on research and life, once in a while.

Sometimes a question needs data we can’t find, so we run the evaluation ourselves. Those results live on the leaderboards, with the code and data to check them.

Who writes it

The site is run by Harit Vishwakarma, who works on uncertainty in machine learning. Guest posts are welcome; see how to contribute. If something is wrong, open an issue or a pull request, and the fix is noted at the bottom of the post.

The name comes from Bayesian statistics. The posterior is what you believe after weighing the evidence, and the logo shows a wide prior narrowing into one.