Turning Karpathy's Autoregressive Baby GPT into Diffusion GPT Step by Step
ash_at_hny Sunday, February 01, 2026
Summary
The article provides a detailed explanation of discrete diffusion models, a powerful class of generative models that can be used for various tasks such as image synthesis, text generation, and molecular design. It covers the mathematical formulation, training, and applications of these models.
1
0
Summary
colab.research.google.com