Learning and memory operations in neural circuits are believed to involve molecular cascades of synaptic and nonsynaptic changes that lead to a diverse repertoire of dynamical phenomena at higher levels of processing. learning to readily demonstrate a spike-timing dependence, stably return to a set-point over long time scales, and remain competitive despite this stability. Beyond unsupervised learning, linking the traces with an external plasticity-modulating signal enables spike-based reinforcement learning. At the postsynaptic neuron, the traces are represented by an activity-dependent ion channel that is shown to regulate the input received by a postsynaptic cell and generate intrinsic graded persistent firing levels. We show how spike-based Hebbian-Bayesian learning can be performed in a simulated inference task using integrate-and-fire (IAF) neurons that are Poisson-firing and background-driven, similar to the preferred regime of cortical neurons. Our results support the view that neurons can represent information in the form of probability distributions, and that probabilistic inference could be a functional by-product of coupled synaptic and nonsynaptic mechanisms operating over several timescales. The model provides a biophysical realization of Bayesian computation by reconciling several observed neural phenomena whose functional effects are only partially understood in concert. presynaptic minicolumns = animal) realized by its observed attributes (e.g., = shape, color, or size). By assuming 697761-98-1 conditional and unconditional independence between as a discrete coded or as an interval coded continuous variable (e.g., = blue, yellow, or pink for = color), a modular network topology follows: minicolumns are distributed into each of hypercolumns (Figure ?(Figure1A).1A). Here, represents relative activity or uncertainty of the attribute value = 1 indicates that attribute value was observed with maximal certainty. Equation 3 may instead be equivalently expressed as a sum of logarithms by: Figure 1 Reconciling neuronal 697761-98-1 and Rabbit Polyclonal to PAK2 (phospho-Ser197) probabilistic spaces using the spike-based BCPNN architecture for a postsynaptic minicolumn with activity = 5 hypercolumns each containing = 4 minicolumns that laterally … can be calculated by iterating over the set of possible conditioning attribute values = for with weight and bias update equations (Figure ?(Figure1B1B): by using an exponential transfer function since = log and as models of the incoming synaptic strength and excitability of a neuron. In the case where multiple synaptic boutons from a pre- to postsynaptic target neuron exist, they are represented here as a single synapse. Probabilistic inference performed with local synaptic traces Spike-based BCPNN is based on memory traces implemented as exponentially weighted moving averages (EWMAs) (Roberts, 1959) of spikes, which were used to estimate as defined above (Equation 5). Temporal smoothing corresponds to integration of neural activity by molecular processes and enables manipulation of 697761-98-1 these traces; it is a technique commonly implemented in synapse (Kempter et al., 1999) and neuron (Gerstner, 1995) models. EWMAs can ensure newly presented evidence is prioritized over previously learned patterns because as old memories decay, they are gradually replaced by more recent ones. The dynamics governing the differential equations of the learning rule with two input spike trains, from presynaptic neuron and from postsynaptic neuron pre- (ACD, red) and postsynaptic (ACD, blue) neuron spike trains are presented as arbitrary example input patterns. Each subsequent row (BCD) … The and traces had the fastest dynamics (Figure ?(Figure2B),2B), and were defined as 5C100 ms to match rapid Ca2+ influx via NMDA receptors or voltage-gated Ca2+ channels (Lisman, 1989; Bliss and Collingridge, 1993). These events initiate synaptic plasticity and can determine the time scale of the coincidence detection window for LTP induction (Markram et al., 1997). We assumed that each neuron could maximally fire at ms, normalizing each spike by meant that it contributed an appropriate proportion of overall probability in a given unit of time by making the underlying trace 1. This established a linear transformation between probability space {?, 1 and neuronal spike rate ? spike rate ?, traces were passed on to the or eligibility traces (Klopf, 1972), which evolved according to (Figure ?(Figure2C2C): traces. Eligibility traces have been used extensively to simulate delayed reward paradigms in previous models (Florian, 2007; Izhikevich, 2007), and are viewed as a potential neural mechanism underlying reinforcement learning (Pawlak et al., 2010). They enabled simultaneous pre-post spiking to trigger.