Postgraduate / Master's · Artificial Intelligence
Explainable AI
What is it
Explainable AI (XAI) studies techniques for making the decisions of complex AI models, particularly deep neural networks, interpretable to humans — covering methods like feature attribution, saliency maps, and inherently interpretable model design, along with the theoretical trade-offs between accuracy and interpretability.
Why it matters
As AI systems are deployed in high-stakes domains like healthcare, finance, and criminal justice, the ability to explain why a model made a particular decision is often a legal and ethical requirement, not just a nice-to-have, making explainability a genuinely central research problem rather than a peripheral concern.
Exam tip
When evaluating an explainability method, distinguish carefully between explanations that are faithful (accurately reflect what the model actually computed) and those that are merely plausible (sound reasonable to a human but may not reflect the model's true reasoning) — this distinction is one of the most important and frequently overlooked issues in the XAI literature.
Related topics
Want help mastering Explainable AI?
Tell us about the student's goals and confidence — we'll design a personalised plan.