Prehistory — From Neural Networks to the TransformerKunihiko Fukushima
Fukushima publishes the Neocognitron
AIMade by AIWritten by AI (Claude) from the source below. Check anything important against the source.How it's made
Published the Neocognitron — a hierarchical pattern-recognition neural network invariant to shifts in position
Why it matters
Introduced local receptive fields and a hierarchy of simple and complex cells, becoming the direct predecessor of the CNN architecture
Source — Primary sources first; secondary reports are labelled “Press”
Paperlink.springer.comhttps://link.springer.com/article/10.1007/BF00344251Added in Week 35, 2026 · See that week
On the map — ⓪ Prehistory line
- Previous stopWerbos's Harvard PhD thesis describes backpropagation for multilayer networks
- This recordNeocognitron
- Next stopBackpropagation