Loading timeline…
19821980s
Hopfield Networks
A recurrent neural network serving as a content-addressable memory system.
Why It Was Important
John Hopfield brought concepts from statistical physics (spin glasses) into neural networks. His model showed how a network could store patterns as local minima in an 'energy landscape' and retrieve the full pattern even from corrupted or incomplete inputs.
Who Invented It
John Hopfield
American physicist bridging physics, biology, and computing.
Applications
- Associative Memory
- Pattern Restoration
- Optimization (Traveling Salesman)
Key Papers
- Neural Networks and Physical Systems with Emergent Collective Computational Abilities
John J. Hopfield · Proceedings of the National Academy of Sciences · 1982
Videos
20: Hopfield Networks - Intro to Neural Computation
MIT OpenCourseWare