anford’s CME295: Transformers & LLMs course provides a deep and structured exploration of modern Large Language Models and their underlying Transformer architecture. The series covers a wide range of topics, starting with the fundamentals of Transformers, their design principles, and how they power state-of-the-art LLMs.
The lectures guide learners through LLM evaluation, teaching methods to measure performance, alignment, and reliability. Advanced topics include LLM reasoning, examining how these models make decisions, and agentic LLMs, exploring the possibilities and implications of models capable of independent task execution. Learners also gain hands-on knowledge of LLM tuning, including fine-tuning techniques to adapt models for specific tasks and applications.
