Spike Timing-Dependent Plasticity

Spike timing-dependent plasticity (STDP) is a Hebbian learning rule in which the magnitude and sign of synaptic modification depend on the precise millisecond-scale timing of pre- and postsynaptic action potentials. When a presynaptic spike precedes a postsynaptic spike by 10–30 milliseconds, the synapse is potentiated; when the order is reversed, it is depressed. This temporal asymmetry encodes causality at the synaptic level: connections that predict postsynaptic firing are strengthened, connections that follow it are weakened.

STDP transforms synaptic plasticity from a correlation-based mechanism into a prediction-based mechanism. It provides the biophysical substrate for learning temporal sequences, detecting causal structure, and stabilizing temporal codes against noise. The temporal window of STDP — tens of milliseconds — matches the timescale of natural sensorimotor contingencies, suggesting that the rule is tuned to the causal structure of embodied interaction. STDP also connects to reward prediction error: dopaminergic signals can gate or modulate STDP, converting local timing rules into global reinforcement learning.\n\n