| Summary: | Part of hippocampal or cortical plasticity is characterized by synaptic modifications as a function of the joint activity of the pre- and postsynaptic neurons. Whether those changes strongly depends on the exact timing, or more on the average firing rates, is still a matter of debate and may vary from areas to areas. However, it has been robustly observed both in vitro and in vivo, that plasticity itself slowly adapts as a function of the dynamic context, and this phenomena is commonly referred to as metaplasticity. An alternative concept considers the regulation of groups of synapses directly to approach a given average firing rate. Then, the change in the strength of a particular synapse of the group (e.g., due to Hebbian learning) will affect others' strengths, which has been coined as heterosynaptic plasticity. Classically, Hebbian synaptic plasticity is paired in neuron network models with such mechanisms so as to stabilize the activity and/or the weight structure. Here, we present a review of various concepts from heterosynaptic plasticity to metaplasticity that have been applied to several spiking models of plasticity, either in hippocampus or in cortex, and how they compete with classical Hebbian learning. We list and discuss the different approaches that are nowadays used by state of the art models of plasticity for incorporating those concepts. Making the point that metaplasticity is an ubiquitous mechanism promoting the stability of neural function over multiple timescales and acting on top of classical Hebbian learning, we stress the need for incorporating it as a key element in the framework of plasticity models.
|
|---|
