LLM Foundations
Understand model behavior, prompts, context limits and output quality.
Choose a lesson for its Quick Review. The questions and answers are being prepared.
Core Concepts
What Is an LLM?
Distinguish language-model capabilities from verified knowledge and reasoning.
Tokens and Tokenization
Understand why text is broken into tokens and how usage is counted.
Context Windows and Missing Information
Recognize context limits, truncation and loss of important details.
Instructions and Message Roles
Explore how different instruction sources affect model behavior.
Prompts and Model Behavior
Sampling and Response Variability
Explain why repeated prompts may produce different valid responses.
Clear Prompts, Constraints and Examples
Construct precise requests with context, goals and examples.
Structured Outputs and Validation
Check whether generated JSON, tables or specified formats meet requirements.
Hallucinations and Uncertainty
Recognize unsupported claims and appropriately expressed uncertainty.
Evaluation and Diagnosis
Grounding Answers in Evidence
Use reference material and verification to support AI-generated answers.
Comparing Models Fairly
Compare outputs using consistent inputs and stated evaluation criteria.
Quality Rubrics and Test Cases
Build repeatable checks for accuracy, relevance, clarity and safety.
Investigating a Poor AI Answer
Trace prompt problems, context gaps and incorrect assumptions.