How the battle works
The Minotaur is made of every idea in this course: loss functions and learning rates, neurons and backpropagation, ablations, error bars and LaTeX. Questions are drawn at random from a pool that spans all eight chambers, and some will ask you to calculate.
- Land 8 correct answers to bring the Minotaur down.
- Every wrong answer costs one of your 3 lives. Lose them all and you retreat, but you can always return.
- There are no retries within a question. Think, then strike.
- Victory earns a large XP bounty, the Minotaur Slayer badge, the legendary Ariadne's Thread card and your certificate of completion. Win without taking a single wound for a secret extra reward.
After the labyrinth
Whatever happens in the arena, look back at how far the thread has come. You started by fitting a line with your hands. Since then you've derived gradient descent and proved when it converges. You've trained networks in your browser and computed their gradients by hand. You've dissected papers, designed rigorous experiments and written them up in LaTeX. That's the core loop of an AI research scientist: learn the maths, run the experiment, write it up honestly, repeat.
The labyrinth goes deeper. More courses are on their way, including transformers, reinforcement learning, generative models and the full research workflow. Until then, the Researcher's Toolkit and your lab notebook are waiting.