The LLM Visualization project displays a 3D model of a GPT-style network. This is the network topology used in OpenAI's GPT-2, GPT-3 (and potentially GPT-4). The first demonstration model shown is a small network that sorts a small list consisting of letters A, B, and C. This is a demonstration example model from Andrej Karpathy's minGPT implementation. The renderer can also visualize networks of any size and works with smaller GPT2 sizes, though the weights are not downloaded (it has hundreds of MB). The CPU Simulation project runs 2D digital circuits with a full editor. The intention is to add some exercises demonstrating things such as: building a simple RISC-V CPU; how components are built down to the gate level: instruction decoding, ALU, addition, etc.; more advanced CPU concepts like different levels of pipelines, caches, etc.