Explainable AI: Unlocking the Black Box of Machine Learning

About this course

Artificial Intelligence is making more decisions than ever — approving loans, diagnosing diseases, moderating content, and powering the chatbots millions of people talk to every day. But there's a problem: most of these systems are black boxes. We know what they decide, but rarely why.

This course is your complete guide to Explainable AI (XAI) — the field dedicated to opening up that black box and building AI systems people can actually trust. This course contains the use of artificial intelligence.

You'll start with the theoretical foundations of Explainable AI, understanding why interpretability matters for trust, safety, and responsible AI adoption. From there, you'll dive into interpretable machine learning, learning the core techniques — both model-based and post-hoc interpretability — used to explain how ML models arrive at their decisions.

Next, we go deeper into deep learning explainability, unpacking how neural networks and recurrent neural networks (RNNs) process information internally, and what makes them so notoriously difficult to interpret.

From there, the course tackles the architecture powering nearly all of today's cutting-edge AI: the Transformer. You'll learn how self-attention and multi-head attention work, why they revolutionized modern AI, and how they can be interpreted to reveal what a model is actually "paying attention to."

With that foundation, you'll move into the world of Large Language Models (LLMs) — what they are, how they work, and why their inner workings remain one of AI's biggest mysteries. You'll explore the evolving landscape of LLMs through a comparative lens, examine prompt engineering as a tool for interpretability, and work through a real-world case analysis comparing ChatGPT and Claude.

Read the full description on Udemy on Udemy (opens in a new tab)

Prices, discounts and availability are set by Udemy. We may earn a commission when you purchase through links on this site.