AI curriculum
Course outline for How Large Language Models Work
Unexpected model outputs are easier to investigate when their mechanisms are understood. Explore language model internals to interpret context, generation, and resource trade-offs.
About How Large Language Models Work
Unexpected model outputs are easier to investigate when their mechanisms are understood. Explore language model internals to interpret context, generation, and resource trade-offs.
How Large Language Models Work Course Objectives
- Explain tokenization, embeddings, and attention.
- Compare context and sampling settings.
- Describe inference trade-offs for reasoning and multimodal tasks.
Pre-requisites
- Basic familiarity with generative AI.
- Comfort reading simple numerical examples.
Lab Setup
- Computer with a browser and a notebook or text editor.
- Supplied inference examples or a small local model; specialized hardware is optional.
Detailed Course Outline
Proposed modulesModule 1: Representing language
- Tokenization
- Embeddings
- Transformer architecture
- Attention
Practical outcome: Trace how text becomes a model input.
Module 2: Managing inference
- Context windows
- Inference
- KV cache
Practical outcome: Estimate how context affects inference resources.
Module 3: Generation behavior
- Sampling
- Reasoning models
- Multimodal models
Practical outcome: Compare outputs under different generation conditions.
Practical exercise
- Analyze a set of model outputs and explain likely context and sampling effects.
How we train
Contact us for full course details, including duration, delivery options and lab requirements.
Ask about exercises, instructor feedback and the prior knowledge you need. Tool-specific courses marked provisional may become modules in a broader course.
Discuss your learning goals