Posts

A deep dive into building a Retrieval-Augmented Generation system in Rust, from understanding the fundamentals to measuring results.

Establishing baselines and exploring learning rate sensitivity - the first steps in neural network experimentation.

Helping elves fit oddly-shaped presents under Christmas trees before the deadline.

Counting paths through a device network to help repair the factory's toroidal reactor.

Toggling indicator lights to activate factory machines with the minimum number of button presses.

Finding the largest rectangle using known corners in a movie theater floor grid.

Connecting junction boxes to form electrical circuits with their nearest neighbours.

Tracing tachyon beams through a manifold and counting all possible timelines.

Helping some cephalopods with their maths homework after falling into a garbage chute.

Optimising the elves stock management of fresh ingredients.

Helping the elves manage paper rolls with accessible neighbour counting and iterative removal.

Day 3 brings battery banks and joltage calculations to restore emergency power to the escalator.

Day 2 brings product ID validation puzzles and some surprising lessons about Python's string counting behaviour.

It's December, which means it's time to take a break from all the usual projects; it's time for the Advent of Code!

Testing the neural network implementation training on logic gates, visualising decision boundaries, and understanding what the network actually learned.

Learning how neural networks work by implementing one from scratch. Part 1 covers the core building blocks: matrices, activations, and backpropagation.

Taking the first steps into machine learning