Hebbian Learning

What is Hebbian learning?

Hebbian learning describes how connections between neurons can change according to their related activity. Named after psychologist Donald Hebb, it is often summarized as “neurons that fire together, wire together.” The phrase captures an idea about learning through co-activity. It leaves out much of the biology.

How does Hebbian learning work?

The core idea is also called Hebb’s rule or Hebb’s law. These names refer to the proposed activity-dependent strengthening principle, not a claim that every connection changes in the same way.

In the classic formulation, repeated activity in one neuron that helps activate another can strengthen the connection between them. This offers a way to think about how experience changes neural connections and how associations can be learned.

The timing matters. In Bi and Poo’s 1998 study of cultured rat hippocampal neurons, different timing relationships between spikes produced strengthening or weakening. This is an example of spike-timing-dependent plasticity.

Hebbian learning, memory, and neural networks

Hebbian ideas help explain how activity can shape connections involved in learning and memory. They also inspire computational learning rules in artificial neural networks, where a connection's weight changes with the activity of the units it connects.

Long-term potentiation is a lasting increase in synaptic strength; it is not another name for every form of Hebbian learning. Nor does Hebbian learning explain every form of memory or plasticity. The slogan is a useful starting point, but “together” does not mean that any two active neurons will automatically build a stronger connection.