Joint Embedding Predictive Architecture (JEPA) is an alternative way to train AI models proposed by Yann LeCun. We’ve been calling the main graphic the Web of AI - it follows the path to the current mainstream approach to AI, LeCun’s alternative path to JEPA, and really nicely shows how discriminative, generative, and joint embedding approaches to AI fit together. The bottom of the poster includes visual summaries of the models we covered in this video - J-VEPA, VL-JEPA, and LeWorldModel.

Download is a high quality pdf.