Training neural networks
How to make neural networks output useful things.
Runs from Jan 9th - 30th 2027.
Facilitators
What it covers
-
Introduction to Neural Networks & inference
Overview of weights, activations, biases, and layers (embeddings, attention and its variants, fully connected layers, output).
-
Backpropagation
Intuitions behind the algorithm, how it works and why it works.
-
Optimisers & Loss functions
-
Types of training
Cover Pre-training, Supervised FineTuning, Post-training, Reinforcement Learning and their purposes.
-
Training data & Evaluations
Signups aren't open yet — express interest and we'll email you when your place can be confirmed.
Express interest