All atlases/Generative AI Fundamentals

AI Atlas

Generative AI Fundamentals

A 10-week introductory course that provides a comprehensive foundation in generative artificial intelligence. Students will learn core concepts, techniques, and applications of generative models including large language models, diffusion models, and GANs. The course covers both theoretical foundations and practical implementation, with emphasis on understanding how these models work, their capabilities and limitations, and how to apply them responsibly. Topics progress from basic generative modeling principles to advanced applications including text generation, image synthesis, and multimodal AI systems.

10 concepts20 checks30 cards

Next: Generative Models and Probability

Atlas map

Foundations of Generative AI

Recommended next1. Generative Models and Probability

← Swipe to explore the full concept map →

current pathprerequisite
Generative Models and Probability appears earlier in the syllabus and supports Autoregressive Generation Fundamentals.Autoregressive Generation Fundamentals appears earlier in the syllabus and supports VAE Architecture and Training.VAE Architecture and Training appears earlier in the syllabus and supports GAN Fundamentals and Training Dynamics.GAN Fundamentals and Training Dynamics appears earlier in the syllabus and supports Modern Diffusion Architectures and Sampling.Modern Diffusion Architectures and Sampling appears earlier in the syllabus and supports Transformer Architecture for Generation.Transformer Architecture for Generation appears earlier in the syllabus and supports Prompt Design and Optimization.Prompt Design and Optimization appears earlier in the syllabus and supports Fine-tuning Fundamentals and Techniques.Fine-tuning Fundamentals and Techniques appears earlier in the syllabus and supports Multimodal Model Architectures.Multimodal Model Architectures appears earlier in the syllabus and supports Evaluating Generative Models.prerequisite

prerequisite relationship

Generative Models and Probability -> Autoregressive Generation Fundamentals

Generative Models and Probability appears earlier in the syllabus and supports Autoregressive Generation Fundamentals.

ready

Recommended next

Generative Models and Probability

This lesson introduces the fundamental concept of generative models as systems that learn to produce new data samples from underlying distributions.

Step 1 / 5

Generative Models and Probability: the core idea

This lesson introduces the fundamental concept of generative models as systems that learn to produce new data samples from underlying distributions. The key thing to notice is: Joint probability distribution p(x). A useful example is Google Smart Compose autocompleting e-mail by modeling p(next word | previous words). Do not treat this as a vocabulary item; the point is to use it to reason about a new situation.

Where would Generative Models and Probability show up in an everyday decision or news headline?

Look for the hidden relationship in the example: Google Smart Compose autocompleting e-mail by modeling p(next word | previous words).

Generative Models and Probability: the core idea