Shon Czinner’s Blog – Home

The 90-year-old idea behind JEPA models: Canonical Correlation Analysis (CCA)

Embedding prediction

ai

jepa

Harold Hotelling’s 1936 Canonical Correlation Analysis (CCA) forms the theoretical and intuitive foundation for modern embedding prediction techniques, including JEPA models.

A Small JEPA Word Embedding Model

ai

code

jepa

After my prior blog post about SIGReg, I figured I’d train a small Joint-Embedding Predictive Architecture (JEPA) model to demonstrate it.

Sketched Isotropic Gaussian Regularization (SIGReg) Explained

Gaussianizing to Prevent Latent Collapse in JEPA Models

ai

code

jepa

The paper “LeJEPA: Provable and Scalable Self-Supervised Learning Without the Heuristics”(Balestriero and LeCun 2025) proposed an interesting method of regularizing latent…

Motivation To Write

meta

I plan to write about AI, statistics, quantitative finance, economics, and more.