How Tiny Neural Networks Represent Basic Functions | Towards Data Science

A gentle introduction to mechanistic interpretability through simple algorithmic examples