John Hopfield

John Hopfield is an American physicist whose work connected ideas from physics and biology to computation. He developed the Hopfield network, a recurrent neural network, and shared the 2024 Nobel Prize in Physics.

Hopfield network: A recurrent neural network that stores patterns as stable states and retrieves them from incomplete or noisy inputs. Hopfield introduced this network in 1982 as a model of associative memory.
Energy function: A scalar function that assigns an energy value to each possible state of a system. Hopfield networks update neurons so their energy decreases until a stable state is reached.
Condensed matter physics: The branch of physics that studies large collections of atoms, molecules, and other interacting particles. Hopfield’s early research in this field supplied tools for modeling collective neural behavior.
Geoffrey Hinton: A computer scientist and cognitive psychologist known for foundational work in neural networks and deep learning. Hinton shared the 2024 Nobel Prize in Physics with Hopfield for foundational machine-learning work.
Boltzmann machine: A stochastic recurrent neural network that learns probability distributions using an energy-based model. It extends Hopfield’s energy-based approach by introducing probabilistic neuron states.
Associative memory: Memory that retrieves a stored item from partial or related cues rather than an exact address. The network models recall by settling into a stored pattern from a partial cue.
Hebbian theory: A learning principle in which connections strengthen when their neurons are active together. Hopfield used Hebbian-style weights to encode patterns in the network.
Hopfield's 1982 paper: John Hopfield’s 1982 paper, “Neural networks and physical systems with emergent collective computational abilities.” It introduced the network and made its physical interpretation explicit.
David Tank: An American physicist and neuroscientist who studies neural circuits and computation. Tank collaborated with Hopfield on experimental and theoretical work in neural computation.
Machine learning: Methods that let computer systems improve performance by learning patterns from data. Hopfield’s network became a foundational model in the history of machine learning.

Linked from 1 page