AI Foundationspredict · compress · act

concepts → Architecture

Architecture

JEPA

also called joint embedding predictive architecture

frontieras of 2026-091 connectiondraft

predict in a learned abstract space, not pixel by pixel

Where it sits

The prism has six jobs across and eight layers down. Its primary cell islearn × L1. Hatched cells cannot exist — a GPU does not learn, an institution does not infer.

What must come first — and what it unlocks

Left to right is reading order, derived from the prerequisite_of edges. Nothing here is hand-ordered: the diagram is the graph.

nothing comes first — this is a starting pointJEPAnothing depends on it yet — a leaf in the reading order

What kind of thing it is — and what it is made of

It is part of: World model.

Where to read it

The chapter that introduces it, and any chapter that uses it again.

42Models of the Worldact 9 · The Frontier

Where it comes from

Every connection

All 1 edges touching this node, grouped by relation family — the sections above are highlights from this list. Colours match the relation families inthe atlas.

Structure · 1
is part ofWorld modelArchitecture

This page is a projection of one node in src/data/concepts.ts. It has no prose file of its own — 230 declared edges produce all 236 of these pages. Edit an edge and both endpoints change.