MATH · IN · MODELS

JEPA (Joint Embedding Predictive Architecture)

Independent / academic

Structures found in this family (5)

Linear Direction, Linear Separability, Platonic Representation Hypothesis, Linear Subspace, Circle

By model (4)

ViT-S/16 (I-JEPA objective, independently trained per view: smallNORB/nuScenes/ImageNet-1k)
Platonic Representation Hypothesis, Linear Subspace
V-JEPA 2
Linear Direction, Circle
V-JEPA (video joint-embedding predictive architecture)
Linear Separability
V-JEPA 2.1
Linear Direction

Observations (5)

A directional-coherence score separates V-JEPA from cosine-similar rivals
Inverting video diffusion transformers makes physical plausibility linearly decodable
Two independently-trained JEPA world models converge to a linear isomorphism
Video world models encode motion direction as a circular population code
Video foundation models vary in physics decodability, V-JEPA strongest

Papers

Latent Video Prediction Learns Better World Models (2026), The Invisible Hand of Physics: When Video Diffusion Models Know More Than They Show (2026), Social-JEPA: Emergent Geometric Isomorphism in Independently Trained World Models (2026), Interpreting Physics in Video World Models (2026), Do Video Foundation Models Understand Intuitive Physics? A Layerwise Probing Analysis (2026)