AI Foundationspredict · compress · act

concepts → Result

Result

Chinchilla

timeless2 connectionsdraft

for a fixed budget, most big models of 2020 were too large and under-trained

Where it sits

The prism has six jobs across and eight layers down. Its primary cell islearn × L4. 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.

Compute-optimalChinchillanothing depends on it yet — a leaf in the reading order

Before it: Compute-optimal

Where to read it

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

27More Is Differentact 6 · The Attention Revolution

Where it comes from

paperTraining Compute-Optimal Large Language ModelsJordan Hoffmann, et al. · 2022

Every connection

All 2 edges touching this node, grouped by relation family — the sections above are highlights from this list. Colours match the relation families inthe atlas.

Order · 1
requiresCompute-optimalProperty
Lineage · 1
improves onScaling lawResult

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.