Subtractive, divisive and non-monotonic gain control in feedforward nets linearized by noise and delays

Abstract

The control of input-to-output mappings, or gain control, is one of the main strategies used by neural networks for the processing and gating of information. Using a spiking neural network model, we studied the gain control induced by a form of inhibitory feedforward circuitry—also known as “open-loop feedback”—, which has been experimentally observed in a cerebellum-like structure in weakly electric fish. We found, both analytically and numerically, that this network displays three different regimes of gain control: subtractive, divisive, and non-monotonic. Subtractive gain control was obtained when noise is very low in the network. Also, it was possible to change from divisive to non-monotonic gain control by simply modulating the strength of the feedforward inhibition, which may be achieved via long-term synaptic plasticity. The particular case of divisive gain control has been previously observed in vivo in weakly electric fish. These gain control regimes were robust to the presence of temporal delays in the inhibitory feedforward pathway, which were found to linearize the input-to-output mappings (or f-I curves) via a novel variability-increasing mechanism. Our findings highlight the feedforward-induced gain control analyzed here as a highly versatile mechanism of information gating in the brain.

Neural activity of heterogeneous inhibitory spiking networks with delay

“…Therefore the neurons are for long time active and unaffected by the other neurons, however when they receive large inhibitory PSPs they remain silent for long periods. It has been shown that heterogeneity and noise can increase the information encoded by a population counteracting the correlation present in neuronal activity [35][36][37][38][39]. However, it remains to clarify how disorder (neural heterogeneity and randomness in the connections) and delay should combine to enhance information encoding.…”

Subthreshold membrane responses underlying sparse spiking to natural vocal signals in auditory cortex

“…We find a major role for inhibition in shaping both the trial‐to‐trial reproducibility of the post‐synaptic response to ongoing natural stimuli, and the magnitude of spike rates – effects that inhibition could manifest by altering the correlation structure of the network. The gain of responses could also be modulated directly by feed‐forward inhibition (Mejias et al ., ). Input–output mappings even in single neurons are not static over time or behavioral conditions (Kozlov & Gentner, ), and our results demonstrate that inhibition is poised to provide flexible control over these response characteristics under natural conditions.…”

A Bayesian Generative Model of Vestibular Afferent Neuron Spiking

“…The most recent observation for a stationary renewal process at any time, the latest available information, is the elapsed time since the most recent event. Many models have been proposed to explain multiplicative gain or sensitivity adjustments and normalization of activity levels across neural populations (Bastian, 1986; Beck, Latham, & Pouget, 2011; Capaday, 2002; Carandini & Heeger, 2012; Eliasmith & Martens, 2011; Louie, Khaw, & Glimcher, 2013; Mejias, Payeur, Selin, Maler, & Longtin, 2014; Nelson, 1994; Olsen, Bhandawat, & Wilson, 2010; Silver, 2010), and we will not consider possible mechanisms for these operations in the vestibular system beyond noting that it is widely accepted that neurons are capable of such computations. The key additional computational capability that neurons would require to implement dynamical Bayesian inference in the vestibular system is the ability to compute parameter likelihoods given the elapsed time since the most recent event.…”