definition

JEPA Predicts Abstract Representations Not Pixels

Joint-Embedding Predictive Architecture (JEPA) trains by encoding both a full input and a corrupted version, then predicting the representation of the full input from the corrupted one in abstract representation space, rather than reconstructing pixels; this lets the encoder discard unpredictable noise and preserve only predictable, modelable structure.

definitionpending

Speaker

Yann LeCun

Evidence Quote

“it allows the system to essentially learn an abstract representation of the world where what can be modeled and predicted is preserved and the rest is viewed as noise and eliminated by the encoder.”

Source

Research Pioneer— Lex Fridman Podcast
Created: 6/13/2026, 3:30:53 AM

Your notes

Loading notes...