Large Language Models (Advanced Module) (2024/2025)

In this course, we will explore the principles, techniques, and applications of LLMs, which are rapidly transforming industries and research fields through their ability to understand and generate human-like text. We will start with fundamental concepts such as autoregressive sequence modeling and neural network architectures and advances toward complex techniques in training, evaluating, and optimizing LLMs. The course culminates in large-scale optimization techniques for training large neural networks on massive text data. As an overarching goal, you will develop the expertise to design, train, evaluate, and optimize large language models. Moreover, through individual and team-based programming assignments, you will apply the presented techniques to solve practical real-world problems. In particular, you will be able to critically assess their societal impact.