MLx Cases 2026

In the Supervised Image Classification module, participants will delve into the construction and application of Convolutional Neural Networks (CNNs) for image classification tasks. The unit emphasizes understanding CNN components, training and evaluation methods, and the significance of activation functions and optimization techniques. Learners will engage in practical exercises to build and refine their own CNN models, gaining the skills necessary to address complex image classification challenges. By the end of the module, participants will be prepared to implement CNNs effectively in various real-world scenarios requiring image analysis.

1. 1. Building a CNN for Supervised Image Classification

This learning unit offers a practical introduction to diffusion models, emphasizing their construction and application in generative modeling. Participants will learn to build a simple diffusion model from scratch, gaining insights into essential components of state-of-the-art models. The unit covers noise addition, the reverse denoising process, and data generation from noise. Through hands-on coding exercises, learners will implement a diffusion model, investigate the effects of varying noise levels, and analyze how conditioning influences generated outputs. By the end, participants will have the foundational knowledge to engage with current research in diffusion models.

1. 1. A Practical Introduction to Diffusion Models

In the "Building LLMs" module of the MLx Cases 2026 learning plan, participants will delve into the creation and development of large language models. The module comprises three units: the first unit focuses on building a mini GPT from scratch in Python, covering essential concepts like tokenization and the self-attention mechanism. The second unit guides learners in creating their own AI agents by integrating LLMs with various tools, emphasizing core components and practical applications. The final unit explores the development of AI agents that can reason, act, and evolve, equipping participants with the skills to create intelligent systems for real-world tasks.

1. 1. Building a Mini-GPT from Scratch in Python
2. 2. Your Own AI Agent
3. 3. Building AI Agents that Reason, React and Evolve!

In the "Building With LLMs" unit of the MLx Cases 2026 learning plan, participants will delve into the practical applications of Large Language Models across various domains. The focus will be on building and automating tasks, highlighting how AI can make coding more accessible. Through hands-on exercises, learners will create applications and visualizations from datasets, fostering a mindset of experimentation and iterative improvement. This unit aims to empower participants to harness AI tools effectively, promoting innovation and self-sufficiency in their coding and automation endeavors.

1. 1. You Can Just Build Things (Really)