Sign inSign in
ConceptAI & ML

Joint embedding predictive architecture

A world model architecture that predicts latent representations instead of raw outputs, focusing on latent structure.
as Steven Feng defines itStanford University
Watch it explained

Know first

Named as a prerequisite by the lectures that define this. The number is how many say so. A grey one has no page yet.

Embedding1Encoder1Generative model1Latent representations1Next-token prediction1Predictive models1

Lecturers

2

Each in their own terms, from the easiest level.

  • Yann LeCun

    Harvard University · Research

    An architecture where both inputs are passed through encoders and predictions are made in an abstract representation space rather than pixel space.
    • 32:32Objective-Driven AI: World Models that Learn, Remember, Reason, and Plan

Connections

6

The lectures that define it or assume it, and the concepts next to it.

2 lecturers. Definitions are written by AI from each lecture’s transcript. A time with a play mark starts the lecture at the moment the lecturer says it.