Abstract
A common feature of both learning and development is that the number and strengths of synaptic connections can change dramatically in a relatively short period of time. Precisely how such changes give rise to functioning and flexible circuits remains an outstanding, unanswered question in neuroscience. One idea, which enjoys widespread support, is that correlation-based rules of synaptic modification are responsible for at least some forms of learning and activity-dependent development in the nervous system.1,2 By these rules, correlated activity in presynaptic and postsynaptic cells leads to an increase in synaptic strength, while uncorrelated activity leads to a decrease in strength. This sort of Hebbian synaptic modification is enormously attractive for a number of reasons. It gives neural circuits great flexibility, and suggests a way for them to wire themselves without extensive genetic instructions. It is also backed by experiments in the hippocampus and neocortex that demonstrate that real synapses can actually follow such rules — namely, in long-term potentiation (LTP) and depression (LTD).2