Inhibitory stabilization of the cortical network underlies visual surround suppression

Learning Fixed Points of Recurrent Neural Networks by Reparameterizing the Network Model

In computational neuroscience, recurrent neural networks are widely used to model neural activity and learning. In many studies, fixed points of recurrent neural networks are used to model neural responses to static or slowly changing stimuli, such as visual cortical responses to static visual stimuli. These applications raise the question of how to train the weights in a recurrent neural network to minimize a loss function evaluated on fixed points. In parallel, training fixed points is a central topic in the study of deep equilibrium models in machine learning. A natural approach is to use gradient descent on the Euclidean space of weights. We show that this approach can lead to poor learning performance due in part to singularities that arise in the loss surface. We use a reparameterization of the recurrent network model to derive two alternative learning rules that produce more robust learning dynamics. We demonstrate that these learning rules avoid singularities and learn more effectively than standard gradient descent. The new learning rules can be interpreted as steepest descent and gradient descent, respectively, under a non-Euclidean metric on the space of recurrent weights. Our results question the common, implicit assumption that learning in the brain should be expected to follow the negative Euclidean gradient of synaptic weights.

The mechanics of correlated variability in segregated cortical excitatory subnetworks

Understanding the genesis of shared trial-to-trial variability in neuronal population activity within the sensory cortex is critical to uncovering the biological basis of information processing in the brain. Shared variability is often a reflection of the structure of cortical connectivity since it likely arises, in part, from local circuit inputs. A series of experiments from segregated networks of (excitatory) pyramidal neurons in the mouse primary visual cortex challenge this view. Specifically, the across-network correlations were found to be larger than predicted given the known weak cross-network connectivity. We aim to uncover the circuit mechanisms responsible for these enhanced correlations through biologically motivated cortical circuit models. Our central finding is that coupling each excitatory subpopulation with a specific inhibitory subpopulation provides the most robust network-intrinsic solution in shaping these enhanced correlations. This result argues for the existence of excitatory-inhibitory functional assemblies in early sensory areas which mirror not just response properties but also connectivity between pyramidal cells. Furthermore, our findings provide theoretical support for recent experimental observations showing that cortical inhibition forms structural and functional subnetworks with excitatory cells, in contrast to the classical view that inhibition is a nonspecific blanket suppression of local excitation.

Synapse-type-specific competitive Hebbian learning forms functional recurrent networks

Cortical networks exhibit complex stimulus-response patterns that are based on specific recurrent interactions between neurons. For example, the balance between excitatory and inhibitory currents has been identified as a central component of cortical computations. However, it remains unclear how the required synaptic connectivity can emerge in developing circuits where synapses between excitatory and inhibitory neurons are simultaneously plastic. Using theory and modeling, we propose that a wide range of cortical response properties can arise from a single plasticity paradigm that acts simultaneously at all excitatory and inhibitory connections-Hebbian learning that is stabilized by the synapse-type-specific competition for a limited supply of synaptic resources. In plastic recurrent circuits, this competition enables the formation and decorrelation of inhibition-balanced receptive fields. Networks develop an assembly structure with stronger synaptic connections between similarly tuned excitatory and inhibitory neurons and exhibit response normalization and orientation-specific center-surround suppression, reflecting the stimulus statistics during training. These results demonstrate how neurons can self-organize into functional networks and suggest an essential role for synapse-type-specific competitive learning in the development of cortical circuits.

The stabilized supralinear network accounts for the contrast dependence of visual cortical gamma oscillations

When stimulated, neural populations in the visual cortex exhibit fast rhythmic activity with frequencies in the gamma band (30-80 Hz). The gamma rhythm manifests as a broad resonance peak in the power-spectrum of recorded local field potentials, which exhibits various stimulus dependencies. In particular, in macaque primary visual cortex (V1), the gamma peak frequency increases with increasing stimulus contrast. Moreover, this contrast dependence is local: when contrast varies smoothly over visual space, the gamma peak frequency in each cortical column is controlled by the local contrast in that column's receptive field. No parsimonious mechanistic explanation for these contrast dependencies of V1 gamma oscillations has been proposed. The stabilized supralinear network (SSN) is a mechanistic model of cortical circuits that has accounted for a range of visual cortical response nonlinearities and contextual modulations, as well as their contrast dependence. Here, we begin by showing that a reduced SSN model without retinotopy robustly captures the contrast dependence of gamma peak frequency, and provides a mechanistic explanation for this effect based on the observed non-saturating and supralinear input-output function of V1 neurons. Given this result, the local dependence on contrast can trivially be captured in a retinotopic SSN which however lacks horizontal synaptic connections between its cortical columns. However, long-range horizontal connections in V1 are in fact strong, and underlie contextual modulation effects such as surround suppression. We thus explored whether a retinotopically organized SSN model of V1 with strong excitatory horizontal connections can exhibit both surround suppression and the local contrast dependence of gamma peak frequency. We found that retinotopic SSNs can account for both effects, but only when the horizontal excitatory projections are composed of two components with different patterns of spatial fall-off with distance: a short-range component that only targets the source column, combined with a long-range component that targets columns neighboring the source column. We thus make a specific qualitative prediction for the spatial structure of horizontal connections in macaque V1, consistent with the columnar structure of cortex.

Cortical determinants of loudness perception and auditory hypersensitivity

Parvalbumin-expressing inhibitory neurons (PVNs) stabilize cortical network activity, generate gamma rhythms, and regulate experience-dependent plasticity. Here, we observed that activation or inactivation of PVNs functioned like a volume knob in the mouse auditory cortex (ACtx), turning neural and behavioral classification of sound level up or down over a 20dB range. PVN loudness adjustments were "sticky", such that a single bout of 40Hz PVN stimulation sustainably suppressed ACtx sound responsiveness, potentiated feedforward inhibition, and behaviorally desensitized mice to loudness. Sensory sensitivity is a cardinal feature of autism, aging, and peripheral neuropathy, prompting us to ask whether PVN stimulation can persistently desensitize mice with ACtx hyperactivity, PVN hypofunction, and loudness hypersensitivity triggered by cochlear sensorineural damage. We found that a single 16-minute bout of 40Hz PVN stimulation session restored normal loudness perception for one week, showing that perceptual deficits triggered by irreversible peripheral injuries can be reversed through targeted cortical circuit interventions.

Ex Vivo Cortical Circuits Learn to Predict and Spontaneously Replay Temporal Patterns

It has been proposed that prediction and timing are computational primitives of neocortical microcircuits, specifically, that neural mechanisms are in place to allow neocortical circuits to autonomously learn the temporal structure of external stimuli and generate internal predictions. To test this hypothesis, we trained cortical organotypic slices on two specific temporal patterns using dual-optical stimulation. After 24-hours of training, whole-cell recordings revealed network dynamics consistent with training-specific timed prediction. Unexpectedly, there was replay of the learned temporal structure during spontaneous activity. Furthermore, some neurons exhibited timed prediction errors. Mechanistically our results indicate that learning relied in part on asymmetric connectivity between distinct neuronal ensembles with temporally-ordered activation. These findings further suggest that local cortical microcircuits are intrinsically capable of learning temporal information and generating predictions, and that the learning rules underlying temporal learning and spontaneous replay can be intrinsic to local cortical microcircuits and not necessarily dependent on top-down interactions.

Single-cell optogenetics reveals attenuation-by-suppression in visual cortical neurons

The relationship between neurons' input and spiking output is central to brain computation. Studies in vitro and in anesthetized animals suggest nonlinearities emerge in cells' input-output (activation) functions as network activity increases, yet how neurons transform inputs in vivo has been unclear. Here, we characterize cortical principal neurons' activation functions in awake mice using two-photon optogenetics. We deliver fixed inputs at the soma while neurons' activity varies with sensory stimuli. We find responses to fixed optogenetic input are nearly unchanged as neurons are excited, reflecting a linear response regime above neurons' resting point. In contrast, responses are dramatically attenuated by suppression. This attenuation is a powerful means to filter inputs arriving to suppressed cells, privileging other inputs arriving to excited neurons. These results have two major implications. First, somatic neural activation functions in vivo accord with the activation functions used in recent machine learning systems. Second, neurons' IO functions can filter sensory inputs - not only do sensory stimuli change neurons' spiking outputs, but these changes also affect responses to input, attenuating responses to some inputs while leaving others unchanged.

Top-down modulation in canonical cortical circuits with short-term plasticity

Cortical dynamics and computations are strongly influenced by diverse GABAergic interneurons, including those expressing parvalbumin (PV), somatostatin (SST), and vasoactive intestinal peptide (VIP). Together with excitatory (E) neurons, they form a canonical microcircuit and exhibit counterintuitive nonlinear phenomena. One instance of such phenomena is response reversal, whereby SST neurons show opposite responses to top-down modulation via VIP depending on the presence of bottom-up sensory input, indicating that the network may function in different regimes under different stimulation conditions. Combining analytical and computational approaches, we demonstrate that model networks with multiple interneuron subtypes and experimentally identified short-term plasticity mechanisms can implement response reversal. Surprisingly, despite not directly affecting SST and VIP activity, PV-to-E short-term depression has a decisive impact on SST response reversal. We show how response reversal relates to inhibition stabilization and the paradoxical effect in the presence of several short-term plasticity mechanisms demonstrating that response reversal coincides with a change in the indispensability of SST for network stabilization. In summary, our work suggests a role of short-term plasticity mechanisms in generating nonlinear phenomena in networks with multiple interneuron subtypes and makes several experimentally testable predictions.

A primary sensory cortical interareal feedforward inhibitory circuit for tacto-visual integration

Tactile sensation and vision are often both utilized for the exploration of objects that are within reach though it is not known whether or how these two distinct sensory systems combine such information. Here in mice, we used a combination of stereo photogrammetry for 3D reconstruction of the whisker array, brain-wide anatomical tracing and functional connectivity analysis to explore the possibility of tacto-visual convergence in sensory space and within the circuitry of the primary visual cortex (VISp). Strikingly, we find that stimulation of the contralateral whisker array suppresses visually evoked activity in a tacto-visual sub-region of VISp whose visual space representation closely overlaps with the whisker search space. This suppression is mediated by local fast-spiking interneurons that receive a direct cortico-cortical input predominantly from layer 6 neurons located in the posterior primary somatosensory barrel cortex (SSp-bfd). These data demonstrate functional convergence within and between two primary sensory cortical areas for multisensory object detection and recognition.

Neural correlates of perceptual similarity masking in primate V1

Visual detection is a fundamental natural task. Detection becomes more challenging as the similarity between the target and the background in which it is embedded increases, a phenomenon termed 'similarity masking'. To test the hypothesis that V1 contributes to similarity masking, we used voltage sensitive dye imaging (VSDI) to measure V1 population responses while macaque monkeys performed a detection task under varying levels of target-background similarity. Paradoxically, we find that during an initial transient phase, V1 responses to the target are enhanced, rather than suppressed, by target-background similarity. This effect reverses in the second phase of the response, so that in this phase V1 signals are positively correlated with the behavioral effect of similarity. Finally, we show that a simple model with delayed divisive normalization can qualitatively account for our findings. Overall, our results support the hypothesis that a nonlinear gain control mechanism in V1 contributes to perceptual similarity masking.

Visual surround suppression at the neural and perceptual levels

Surround suppression was initially identified as a phenomenon at the neural level in which stimuli outside the neuron's receptive field alone cannot activate responses but can modulate neural responses to stimuli covered inside the receptive field. Subsequent studies showed that surround suppression is not only a critical property of neurons across species and brain areas but also has been found in visual perceptions. More importantly, surround suppression varies across individuals and shows significant differences between normal controls and patients with certain mental disorders. Here, we combined results from related literature and summarized the findings derived from physiological and psychophysical evidence. We first outline the basic properties of surround suppression in the visual system and perceptions. Then, we mainly summarize the differences in perceptual surround suppression among different human subjects. Our review suggests that there is no consensus regarding whether the strength of perceptual surround suppression could be used as an effective index to distinguish particular populations. Then, we summarized the similar mechanisms for surround suppression and cognitive impairments to further explore the potential clinical applications of surround suppression. A clearer understanding of the mechanisms of surround suppression in neural responses and perceptions is necessary for facilitating its clinical applications.

Response sub-additivity and variability quenching in visual cortex

Sub-additivity and variability are ubiquitous response motifs in the primary visual cortex (V1). Response sub-additivity enables the construction of useful interpretations of the visual environment, whereas response variability indicates the factors that limit the precision with which the brain can do this. There is increasing evidence that experimental manipulations that elicit response sub-additivity often also quench response variability. Here, we provide an overview of these phenomena and suggest that they may have common origins. We discuss empirical findings and recent model-based insights into the functional operations, computational objectives and circuit mechanisms underlying V1 activity. These different modelling approaches all predict that response sub-additivity and variability quenching often co-occur. The phenomenology of these two response motifs, as well as many of the insights obtained about them in V1, generalize to other cortical areas. Thus, the connection between response sub-additivity and variability quenching may be a canonical motif across the cortex.

The influence of cortical activity on perception depends on behavioral state and sensory context

The mechanistic link between neural circuit activity and behavior remains unclear. While manipulating cortical activity can bias certain behaviors and elicit artificial percepts, some tasks can still be solved when cortex is silenced or removed. Here, mice were trained to perform a visual detection task during which we selectively targeted groups of visually responsive and co-tuned neurons in L2/3 of primary visual cortex (V1) for two-photon photostimulation. The influence of photostimulation was conditional on two key factors: the behavioral state of the animal and the contrast of the visual stimulus. The detection of low-contrast stimuli was enhanced by photostimulation, while the detection of high-contrast stimuli was suppressed, but crucially, only when mice were highly engaged in the task. When mice were less engaged, our manipulations of cortical activity had no effect on behavior. The behavioral changes were linked to specific changes in neuronal activity. The responses of non-photostimulated neurons in the local network were also conditional on two factors: their functional similarity to the photostimulated neurons and the contrast of the visual stimulus. Functionally similar neurons were increasingly suppressed by photostimulation with increasing visual stimulus contrast, correlating with the change in behavior. Our results show that the influence of cortical activity on perception is not fixed, but dynamically and contextually modulated by behavioral state, ongoing activity and the routing of information through specific circuits.

Stable fixed points of combinatorial threshold-linear networks

Combinatorial threshold-linear networks (CTLNs) are a special class of recurrent neural networks whose dynamics are tightly controlled by an underlying directed graph. Recurrent networks have long been used as models for associative memory and pattern completion, with stable fixed points playing the role of stored memory patterns in the network. In prior work, we showed that target-free cliques of the graph correspond to stable fixed points of the dynamics, and we conjectured that these are the only stable fixed points possible [1, 2]. In this paper, we prove that the conjecture holds in a variety of special cases, including for networks with very strong inhibition and graphs of size n ≤ 4 . We also provide further evi-dence for the conjecture by showing that sparse graphs and graphs that are nearly cliques can never support stable fixed points. Finally, we translate some results from extremal com-binatorics to obtain an upper bound on the number of stable fixed points of CTLNs in cases where the conjecture holds.

Primate neocortex performs balanced sensory amplification

The sensory cortex amplifies relevant features of external stimuli. This sensitivity and selectivity arise through the transformation of inputs by cortical circuitry. We characterize the circuit mechanisms and dynamics of cortical amplification by making large-scale simultaneous measurements of single cells in awake primates and testing computational models. By comparing network activity in both driven and spontaneous states with models, we identify the circuit as operating in a regime of non-normal balanced amplification. Incoming inputs are strongly but transiently amplified by strong recurrent feedback from the disruption of excitatory-inhibitory balance in the network. Strong inhibition rapidly quenches responses, thereby permitting the tracking of time-varying stimuli.

Learning what matters: Synaptic plasticity with invariance to second-order input correlations

Cortical populations of neurons develop sparse representations adapted to the statistics of the environment. To learn efficient population codes, synaptic plasticity mechanisms must differentiate relevant latent features from spurious input correlations, which are omnipresent in cortical networks. Here, we develop a theory for sparse coding and synaptic plasticity that is invariant to second-order correlations in the input. Going beyond classical Hebbian learning, our learning objective explains the functional form of observed excitatory plasticity mechanisms, showing how Hebbian long-term depression (LTD) cancels the sensitivity to second-order correlations so that receptive fields become aligned with features hidden in higher-order statistics. Invariance to second-order correlations enhances the versatility of biologically realistic learning models, supporting optimal decoding from noisy inputs and sparse population coding from spatially correlated stimuli. In a spiking model with triplet spike-timing-dependent plasticity (STDP), we show that individual neurons can learn localized oriented receptive fields, circumventing the need for input preprocessing, such as whitening, or population-level lateral inhibition. The theory advances our understanding of local unsupervised learning in cortical circuits, offers new interpretations of the Bienenstock-Cooper-Munro and triplet STDP models, and assigns a specific functional role to synaptic LTD mechanisms in pyramidal neurons.

The logic of recurrent circuits in the primary visual cortex

Recurrent cortical activity sculpts visual perception by refining, amplifying or suppressing visual input. However, the rules that govern the influence of recurrent activity remain enigmatic. We used ensemble-specific two-photon optogenetics in the mouse visual cortex to isolate the impact of recurrent activity from external visual input. We found that the spatial arrangement and the visual feature preference of the stimulated ensemble and the neighboring neurons jointly determine the net effect of recurrent activity. Photoactivation of these ensembles drives suppression in all cells beyond 30 µm but uniformly drives activation in closer similarly tuned cells. In nonsimilarly tuned cells, compact, cotuned ensembles drive net suppression, while diffuse, cotuned ensembles drive activation. Computational modeling suggests that highly local recurrent excitatory connectivity and selective convergence onto inhibitory neurons explain these effects. Our findings reveal a straightforward logic in which space and feature preference of cortical ensembles determine their impact on local recurrent activity.

Serotonergic Modulation of the Excitation/Inhibition Balance in the Visual Cortex

Serotonergic neurons constitute one of the main systems of neuromodulators, whose diffuse projections regulate the functions of the cerebral cortex. Serotonin (5-HT) is known to play a crucial role in the differential modulation of cortical activity related to behavioral contexts. Some features of the 5-HT signaling organization suggest its possible participation as a modulator of activity-dependent synaptic changes during the critical period of the primary visual cortex (V1). Cells of the serotonergic system are among the first neurons to differentiate and operate. During postnatal development, ramifications from raphe nuclei become massively distributed in the visual cortical area, remarkably increasing the availability of 5-HT for the regulation of excitatory and inhibitory synaptic activity. A substantial amount of evidence has demonstrated that synaptic plasticity at pyramidal neurons of the superficial layers of V1 critically depends on a fine regulation of the balance between excitation and inhibition (E/I). 5-HT could therefore play an important role in controlling this balance, providing the appropriate excitability conditions that favor synaptic modifications. In order to explore this possibility, the present work used in vitro intracellular electrophysiological recording techniques to study the effects of 5-HT on the E/I balance of V1 layer 2/3 neurons, during the critical period. Serotonergic action on the E/I balance has been analyzed on spontaneous activity, evoked synaptic responses, and long-term depression (LTD). Our results pointed out that the predominant action of 5-HT implies a reduction in the E/I balance. 5-HT promoted LTD at excitatory synapses while blocking it at inhibitory synaptic sites, thus shifting the Hebbian alterations of synaptic strength towards lower levels of E/I balance.

Circuit mechanism for suppression of frontal cortical ignition during NREM sleep

Conscious perception is greatly diminished during sleep, but the underlying circuit mechanism is poorly understood. We show that cortical ignition-a brain process shown to be associated with conscious awareness in humans and non-human primates-is strongly suppressed during non-rapid-eye-movement (NREM) sleep in mice due to reduced cholinergic modulation and rapid inhibition of cortical responses. Brain-wide functional ultrasound imaging and cell-type-specific calcium imaging combined with optogenetics showed that activity propagation from visual to frontal cortex is markedly reduced during NREM sleep due to strong inhibition of frontal pyramidal neurons. Chemogenetic activation and inactivation of basal forebrain cholinergic neurons powerfully increased and decreased visual-to-frontal activity propagation, respectively. Furthermore, although multiple subtypes of dendrite-targeting GABAergic interneurons in the frontal cortex are more active during wakefulness, soma-targeting parvalbumin-expressing interneurons are more active during sleep. Chemogenetic manipulation of parvalbumin interneurons showed that sleep/wake-dependent cortical ignition is strongly modulated by perisomatic inhibition of pyramidal neurons.

Excitation creates a distributed pattern of cortical suppression due to varied recurrent input

Dense local, recurrent connections are a major feature of cortical circuits, yet how they affect neurons' responses has been unclear, with some studies reporting weak recurrent effects, some reporting amplification, and others indicating local suppression. Here, we show that optogenetic input to mouse V1 excitatory neurons generates salt-and-pepper patterns of both excitation and suppression. Responses in individual neurons are not strongly predicted by that neuron's direct input. A balanced-state network model reconciles a set of diverse observations: the observed dynamics, suppressed responses, decoupling of input and output, and long tail of excited responses. The model shows recurrent excitatory-excitatory connections are strong and also variable across neurons. Together, these results demonstrate that excitatory recurrent connections can have major effects on cortical computations by shaping and changing neurons' responses to input.

Mechanisms underlying reshuffling of visual responses by optogenetic stimulation in mice and monkeys

The ability to optogenetically perturb neural circuits opens an unprecedented window into mechanisms governing circuit function. We analyzed and theoretically modeled neuronal responses to visual and optogenetic inputs in mouse and monkey V1. In both species, optogenetic stimulation of excitatory neurons strongly modulated the activity of single neurons yet had weak or no effects on the distribution of firing rates across the population. Thus, the optogenetic inputs reshuffled firing rates across the network. Key statistics of mouse and monkey responses lay on a continuum, with mice/monkeys occupying the low-/high-rate regions, respectively. We show that neuronal reshuffling emerges generically in randomly connected excitatory/inhibitory networks, provided the coupling strength (combination of recurrent coupling and external input) is sufficient that powerful inhibitory feedback cancels the mean optogenetic input. A more realistic model, distinguishing tuned visual vs. untuned optogenetic input in a structured network, reduces the coupling strength needed to explain reshuffling.

Amplitude- and frequency-dependent activation of layer II/III neurons by intracortical microstimulation

Intracortical microstimulation (ICMS) has been used for the development of brain machine interfaces. However, further understanding about the spatiotemporal responses of neurons to different electrical stimulation parameters is necessary to inform the design of optimal therapies. In this study, we employed in vivo electrophysiological recording, two-photon calcium imaging, and electric field simulation to evaluate the acute effect of ICMS on layer II/III neurons. Our results show that stimulation frequency non-linearly modulates neuronal responses, whereas the magnitude of responses is linearly correlated to the electric field strength and stimulation amplitude before reaching a steady state. Temporal dynamics of neurons' responses depends more on stimulation frequency and their distance to the stimulation electrode. In addition, amplitude-dependent post-stimulation suppression was observed within ∼500 μm of the stimulation electrode, as evidenced by both calcium imaging and local field potentials. These findings provide insights for selecting stimulation parameters to achieve desirable spatiotemporal specificity of ICMS.

Lateral inhibition in V1 controls neural & perceptual contrast sensitivity

Lateral inhibition is a central principle for sensory system function. It is thought to operate by the activation of inhibitory neurons that restrict the spatial spread of sensory excitation. Much work on the role of inhibition in sensory systems has focused on visual cortex; however, the neurons, computations, and mechanisms underlying cortical lateral inhibition remain debated, and its importance for visual perception remains unknown. Here, we tested how lateral inhibition from PV or SST neurons in mouse primary visual cortex (V1) modulates neural and perceptual sensitivity to stimulus contrast. Lateral inhibition from PV neurons reduced neural and perceptual sensitivity to visual contrast in a uniform subtractive manner, whereas lateral inhibition from SST neurons more effectively changed the slope (or gain) of neural and perceptual contrast sensitivity. A neural circuit model identified spatially extensive lateral projections from SST neurons as the key factor, and we confirmed this with direct subthreshold measurements of a larger spatial footprint for SST versus PV lateral inhibition. Together, these results define cell-type specific computational roles for lateral inhibition in V1, and establish their unique consequences on sensitivity to contrast, a fundamental aspect of the visual world.

Response retention and apparent motion effect in visual cortex models

Apparent motion is a visual illusion in which stationary stimuli, flashing in distinct spatial locations at certain time intervals, are perceived as one stimulus moving between these locations. In the primary visual cortex, apparent-motion stimuli produce smooth spatio-temporal patterns of activity similar to those produced by continuously moving stimuli. An important prerequisite for producing such activity patterns is prolongation of responses to brief stimuli. Indeed, a brief stimulus can evoke in the visual cortex a long response, outlasting the stimulus by hundreds of milliseconds. Here we use firing-rate based models with simple ring structure, and biologically-detailed conductance-based refractory density (CBRD) model with retinotopic space representation to analyze the response retention and the origin of smooth profiles of activity in response to apparent-motion stimuli. We show that the strength of recurrent connectivity is the major factor that endorses neuronal networks with the ability for response retention. The same strengths of recurrent connections mediate the appearance of bump attractor in the ring models. Factors such as synaptic depression, NMDA receptor mediated currents, and conductances regulating spike adaptation influence response retention, but cannot substitute for the weakness of recurrent connections to reproduce response retention in models with weak connectivity. However, the weakness of lateral recurrent connections can be compensated by layering: in multi-layer models even with weaker connections the activity retains due to its feedforward propagation from layer to layer. Using CBRD model with retinotopic space representation we further show that smooth spatio-temporal profiles of activity in response to apparent-motion stimuli are produced in the models expressing response retention, but not in the models that fail to produce response retention. Together, these results demonstrate a link between response retention and the ability of neuronal networks to generate spatio-temporal patterns of activity, which are compatible with perception of apparent motion.

Adaptive oscillators support Bayesian prediction in temporal processing

Humans excel at predictively synchronizing their behavior with external rhythms, as in dance or music performance. The neural processes underlying rhythmic inferences are debated: whether predictive perception relies on high-level generative models or whether it can readily be implemented locally by hard-coded intrinsic oscillators synchronizing to rhythmic input remains unclear and different underlying computational mechanisms have been proposed. Here we explore human perception for tone sequences with some temporal regularity at varying rates, but with considerable variability. Next, using a dynamical systems perspective, we successfully model the participants behavior using an adaptive frequency oscillator which adjusts its spontaneous frequency based on the rate of stimuli. This model better reflects human behavior than a canonical nonlinear oscillator and a predictive ramping model-both widely used for temporal estimation and prediction-and demonstrate that the classical distinction between absolute and relative computational mechanisms can be unified under this framework. In addition, we show that neural oscillators may constitute hard-coded physiological priors-in a Bayesian sense-that reduce temporal uncertainty and facilitate the predictive processing of noisy rhythms. Together, the results show that adaptive oscillators provide an elegant and biologically plausible means to subserve rhythmic inference, reconciling previously incompatible frameworks for temporal inferential processes.

NEURAL CORRELATES OF PERCEPTUAL SIMILARITY MASKING IN PRIMATE V1

Visual detection is a fundamental natural task. Detection becomes more challenging as the similarity between the target and the background in which it is embedded increases, a phenomenon termed "similarity masking". To test the hypothesis that V1 contributes to similarity masking, we used voltage sensitive dye imaging (VSDI) to measure V1 population responses while macaque monkeys performed a detection task under varying levels of target-background similarity. Paradoxically, we find that during an initial transient phase, V1 responses to the target are enhanced, rather than suppressed, by target-background similarity. This effect reverses in the second phase of the response, so that in this phase V1 signals are positively correlated with the behavioral effect of similarity. Finally, we show that a simple model with delayed divisive normalization can qualitatively account for our findings. Overall, our results support the hypothesis that a nonlinear gain control mechanism in V1 contributes to perceptual similarity masking.

VIP interneurons regulate cortical size tuning and visual perception

Cortical circuit function is regulated by extensively interconnected, diverse populations of GABAergic interneurons that may play key roles in shaping circuit operation according to behavioral context. A specialized population of interneurons that co-express vasoactive intestinal peptides (VIP-INs) are activated during arousal and innervate other INs and pyramidal neurons (PNs). Although state-dependent modulation of VIP-INs has been extensively studied, their role in regulating sensory processing is less well understood. We examined the impact of VIP-INs in the primary visual cortex of awake behaving mice. Loss of VIP-IN activity alters the behavioral state-dependent modulation of somatostatin-expressing INs (SST-INs) but not PNs. In contrast, reduced VIP-IN activity globally disrupts visual feature selectivity for stimulus size. Moreover, the impact of VIP-INs on perceptual behavior varies with context and is more acute for small than large visual cues. VIP-INs thus contribute to both state-dependent modulation of cortical activity and sensory context-dependent perceptual performance.

Generalizing biological surround suppression based on center surround similarity via deep neural network models

Sensory perception is dramatically influenced by the context. Models of contextual neural surround effects in vision have mostly accounted for Primary Visual Cortex (V1) data, via nonlinear computations such as divisive normalization. However, surround effects are not well understood within a hierarchy, for neurons with more complex stimulus selectivity beyond V1. We utilized feedforward deep convolutional neural networks and developed a gradient-based technique to visualize the most suppressive and excitatory surround. We found that deep neural networks exhibited a key signature of surround effects in V1, highlighting center stimuli that visually stand out from the surround and suppressing responses when the surround stimulus is similar to the center. We found that in some neurons, especially in late layers, when the center stimulus was altered, the most suppressive surround surprisingly can follow the change. Through the visualization approach, we generalized previous understanding of surround effects to more complex stimuli, in ways that have not been revealed in visual cortices. In contrast, the suppression based on center surround similarity was not observed in an untrained network. We identified further successes and mismatches of the feedforward CNNs to the biology. Our results provide a testable hypothesis of surround effects in higher visual cortices, and the visualization approach could be adopted in future biological experimental designs.

Propagation of activity through the cortical hierarchy and perception are determined by neural variability

Brains are composed of anatomically and functionally distinct regions performing specialized tasks, but regions do not operate in isolation. Orchestration of complex behaviors requires communication between brain regions, but how neural dynamics are organized to facilitate reliable transmission is not well understood. Here we studied this process directly by generating neural activity that propagates between brain regions and drives behavior, assessing how neural populations in sensory cortex cooperate to transmit information. We achieved this by imaging two densely interconnected regions-the primary and secondary somatosensory cortex (S1 and S2)-in mice while performing two-photon photostimulation of S1 neurons and assigning behavioral salience to the photostimulation. We found that the probability of perception is determined not only by the strength of the photostimulation but also by the variability of S1 neural activity. Therefore, maximizing the signal-to-noise ratio of the stimulus representation in cortex relative to the noise or variability is critical to facilitate activity propagation and perception.

Inhibitory stabilized network behaviour in a balanced neural mass model of a cortical column

Strong inhibitory recurrent connections can reduce the tendency for a neural network to become unstable. This is known as inhibitory stabilization; networks that are unstable in the absence of strong inhibitory feedback because of their unstable excitatory recurrent connections are known as Inhibition Stabilized Networks (ISNs). One of the characteristics of ISNs is their "paradoxical response", where perturbing the inhibitory neurons with additional excitatory input results in a decrease in their activity after a temporal delay instead of increasing their activity. Here, we develop a model of populations of neurons across different layers of cortex. Within each layer, there is one population of inhibitory neurons and one population of excitatory neurons. The connectivity weights across different populations in the model are derived from a synaptic physiology database provided by the Allen Institute. The model shows a gradient of excitation-inhibition balance across different layers in the cortex, where superficial layers are more inhibitory dominated compared to deeper layers. To investigate the presence of ISNs across different layers, we measured the membrane potentials of neural populations in the model after perturbing inhibitory populations. The results show that layer 2/3 in the model does not operate in the ISN regime but layers 4 and 5 do operate in the ISN regime. These results accord with neurophysiological findings that explored the presence of ISNs across different layers in the cortex. The results show that there may be a systematic macroscopic gradient of inhibitory stabilization across different layers in the cortex that depends on the level of excitation-inhibition balance, and that the strength of the paradoxical response increases as the model moves closer to bifurcation points.

Investigating the ability of astrocytes to drive neural network synchrony

Recent experimental works have implicated astrocytes as a significant cell type underlying several neuronal processes in the mammalian brain, from encoding sensory information to neurological disorders. Despite this progress, it is still unclear how astrocytes are communicating with and driving their neuronal neighbors. While previous computational modeling works have helped propose mechanisms responsible for driving these interactions, they have primarily focused on interactions at the synaptic level, with microscale models of calcium dynamics and neurotransmitter diffusion. Since it is computationally infeasible to include the intricate microscale details in a network-scale model, little computational work has been done to understand how astrocytes may be influencing spiking patterns and synchronization of large networks. We overcome this issue by first developing an "effective" astrocyte that can be easily implemented to already established network frameworks. We do this by showing that the astrocyte proximity to a synapse makes synaptic transmission faster, weaker, and less reliable. Thus, our "effective" astrocytes can be incorporated by considering heterogeneous synaptic time constants, which are parametrized only by the degree of astrocytic proximity at that synapse. We then apply our framework to large networks of exponential integrate-and-fire neurons with various spatial structures. Depending on key parameters, such as the number of synapses ensheathed and the strength of this ensheathment, we show that astrocytes can push the network to a synchronous state and exhibit spatially correlated patterns.

Laminar mechanisms of saccadic suppression in primate visual cortex

Saccadic eye movements are known to cause saccadic suppression, a temporary reduction in visual sensitivity and visual cortical firing rates. While saccadic suppression has been well characterized at the level of perception and single neurons, relatively little is known about the visual cortical networks governing this phenomenon. Here we examine the effects of saccadic suppression on distinct neural subpopulations within visual area V4. We find subpopulation-specific differences in the magnitude and timing of peri-saccadic modulation. Input-layer neurons show changes in firing rate and inter-neuronal correlations prior to saccade onset, and putative inhibitory interneurons in the input layer elevate their firing rate during saccades. A computational model of this circuit recapitulates our empirical observations and demonstrates that an input-layer-targeting pathway can initiate saccadic suppression by enhancing local inhibitory activity. Collectively, our results provide a mechanistic understanding of how eye movement signaling interacts with cortical circuitry to enforce visual stability.

Cell-type-specific plasticity of inhibitory interneurons in the rehabilitation of auditory cortex after peripheral damage

Peripheral sensory organ damage leads to compensatory cortical plasticity that is associated with a remarkable recovery of cortical responses to sound. The precise mechanisms that explain how this plasticity is implemented and distributed over a diverse collection of excitatory and inhibitory cortical neurons remain unknown. After noise trauma and persistent peripheral deficits, we found recovered sound-evoked activity in mouse A1 excitatory principal neurons (PNs), parvalbumin- and vasoactive intestinal peptide-expressing neurons (PVs and VIPs), but reduced activity in somatostatin-expressing neurons (SOMs). This cell-type-specific recovery was also associated with cell-type-specific intrinsic plasticity. These findings, along with our computational modelling results, are consistent with the notion that PV plasticity contributes to PN stability, SOM plasticity allows for increased PN and PV activity, and VIP plasticity enables PN and PV recovery by inhibiting SOMs.

Targeting operational regimes of interest in recurrent neural networks

Neural computations emerge from local recurrent neural circuits or computational units such as cortical columns that comprise hundreds to a few thousand neurons. Continuous progress in connectomics, electrophysiology, and calcium imaging require tractable spiking network models that can consistently incorporate new information about the network structure and reproduce the recorded neural activity features. However, for spiking networks, it is challenging to predict which connectivity configurations and neural properties can generate fundamental operational states and specific experimentally reported nonlinear cortical computations. Theoretical descriptions for the computational state of cortical spiking circuits are diverse, including the balanced state where excitatory and inhibitory inputs balance almost perfectly or the inhibition stabilized state (ISN) where the excitatory part of the circuit is unstable. It remains an open question whether these states can co-exist with experimentally reported nonlinear computations and whether they can be recovered in biologically realistic implementations of spiking networks. Here, we show how to identify spiking network connectivity patterns underlying diverse nonlinear computations such as XOR, bistability, inhibitory stabilization, supersaturation, and persistent activity. We establish a mapping between the stabilized supralinear network (SSN) and spiking activity which allows us to pinpoint the location in parameter space where these activity regimes occur. Notably, we find that biologically-sized spiking networks can have irregular asynchronous activity that does not require strong excitation-inhibition balance or large feedforward input and we show that the dynamic firing rate trajectories in spiking networks can be precisely targeted without error-driven training algorithms.

The stabilized supralinear network accounts for the contrast dependence of visual cortical gamma oscillations

When stimulated, neural populations in the visual cortex exhibit fast rhythmic activity with frequencies in the gamma band (30-80 Hz). The gamma rhythm manifests as a broad resonance peak in the power-spectrum of recorded local field potentials, which exhibits various stimulus dependencies. In particular, in macaque primary visual cortex (V1), the gamma peak frequency increases with increasing stimulus contrast. Moreover, this contrast dependence is local: when contrast varies smoothly over visual space, the gamma peak frequency in each cortical column is controlled by the local contrast in that column's receptive field. No parsimonious mechanistic explanation for these contrast dependencies of V1 gamma oscillations has been proposed. The stabilized supralinear network (SSN) is a mechanistic model of cortical circuits that has accounted for a range of visual cortical response nonlinearities and contextual modulations, as well as their contrast dependence. Here, we begin by showing that a reduced SSN model without retinotopy robustly captures the contrast dependence of gamma peak frequency, and provides a mechanistic explanation for this effect based on the observed non-saturating and supralinear input-output function of V1 neurons. Given this result, the local dependence on contrast can trivially be captured in a retinotopic SSN which however lacks horizontal synaptic connections between its cortical columns. However, long-range horizontal connections in V1 are in fact strong, and underlie contextual modulation effects such as surround suppression. We thus explored whether a retinotopically organized SSN model of V1 with strong excitatory horizontal connections can exhibit both surround suppression and the local contrast dependence of gamma peak frequency. We found that retinotopic SSNs can account for both effects, but only when the horizontal excitatory projections are composed of two components with different patterns of spatial fall-off with distance: a short-range component that only targets the source column, combined with a long-range component that targets columns neighboring the source column. We thus make a specific qualitative prediction for the spatial structure of horizontal connections in macaque V1, consistent with the columnar structure of cortex.

Distinct nonlinear spectrotemporal integration in primary and secondary auditory cortices

Animals sense sounds through hierarchical neural pathways that ultimately reach higher-order cortices to extract complex acoustic features, such as vocalizations. Elucidating how spectrotemporal integration varies along the hierarchy from primary to higher-order auditory cortices is a crucial step in understanding this elaborate sensory computation. Here we used two-photon calcium imaging and two-tone stimuli with various frequency-timing combinations to compare spectrotemporal integration between primary (A1) and secondary (A2) auditory cortices in mice. Individual neurons showed mixed supralinear and sublinear integration in a frequency-timing combination-specific manner, and we found unique integration patterns in these two areas. Temporally asymmetric spectrotemporal integration in A1 neurons suggested their roles in discriminating frequency-modulated sweep directions. In contrast, temporally symmetric and coincidence-preferring integration in A2 neurons made them ideal spectral integrators of concurrent multifrequency sounds. Moreover, the ensemble neural activity in A2 was sensitive to two-tone timings, and coincident two-tones evoked distinct ensemble activity patterns from the linear sum of component tones. Together, these results demonstrate distinct roles of A1 and A2 in encoding complex acoustic features, potentially suggesting parallel rather than sequential information extraction between these regions.

Optogenetics reveals paradoxical network stabilizations in hippocampal CA1 and CA3

Recurrent connectivity between excitatory neurons and the strength of feedback from inhibitory neurons are critical determinants of the dynamics and computational properties of neuronal circuits. Toward a better understanding of these circuit properties in regions CA1 and CA3 of the hippocampus, we performed optogenetic manipulations combined with large-scale unit recordings in rats under anesthesia and in quiet waking, using photoinhibition and photoexcitation with different light-sensitive opsins. In both regions, we saw striking paradoxical responses: subsets of cells increased firing during photoinhibition, while other cells decreased firing during photoexcitation. These paradoxical responses were more prominent in CA3 than in CA1, but, notably, CA1 interneurons showed increased firing in response to photoinhibition of CA3. These observations were recapitulated in simulations where we modeled both CA1 and CA3 as inhibition-stabilized networks in which strong recurrent excitation is balanced by feedback inhibition. To directly test the inhibition-stabilized model, we performed large-scale photoinhibition directed at (GAD-Cre) inhibitory cells and found that interneurons in both regions increased firing when photoinhibited, as predicted. Our results highlight the often-paradoxical circuit dynamics that are evidenced during optogenetic manipulations and indicate that, contrary to long-standing dogma, both CA1 and CA3 hippocampal regions display strongly recurrent excitation, which is stabilized through inhibition.

VIP interneurons regulate cortical size tuning and visual perception

Local cortical circuit function is regulated by diverse populations of GABAergic interneurons with distinct properties and extensive interconnectivity. Inhibitory-to-inhibitory interactions between interneuron populations may play key roles in shaping circuit operation according to behavioral context. A specialized population of GABAergic interneurons that co-express vasoactive intestinal peptide (VIP-INs) are activated during arousal and locomotion and innervate other local interneurons and pyramidal neurons. Although modulation of VIP-IN activity by behavioral state has been extensively studied, their role in regulating information processing and selectivity is less well understood. Using a combination of cellular imaging, short and long-term manipulation, and perceptual behavior, we examined the impact of VIP-INs on their synaptic target populations in the primary visual cortex of awake behaving mice. We find that loss of VIP-IN activity alters the behavioral state-dependent modulation of somatostatin-expressing interneurons (SST-INs) but not pyramidal neurons (PNs). In contrast, reduced VIP-IN activity disrupts visual feature selectivity for stimulus size in both populations. Inhibitory-to inhibitory interactions thus directly shape the selectivity of GABAergic interneurons for sensory stimuli. Moreover, the impact of VIP-IN activity on perceptual behavior varies with visual context and is more acute for small than large visual cues. VIP-INs thus contribute to both state-dependent modulation of cortical circuit activity and sensory context-dependent perceptual performance.

The mechanics of correlated variability in segregated cortical excitatory subnetworks

Understanding the genesis of shared trial-to-trial variability in neural activity within sensory cortex is critical to uncovering the biological basis of information processing in the brain. Shared variability is often a reflection of the structure of cortical connectivity since this variability likely arises, in part, from local circuit inputs. A series of experiments from segregated networks of (excitatory) pyramidal neurons in mouse primary visual cortex challenge this view. Specifically, the across-network correlations were found to be larger than predicted given the known weak cross-network connectivity. We aim to uncover the circuit mechanisms responsible for these enhanced correlations through biologically motivated cortical circuit models. Our central finding is that coupling each excitatory subpopulation with a specific inhibitory subpopulation provides the most robust network-intrinsic solution in shaping these enhanced correlations. This result argues for the existence of excitatory-inhibitory functional assemblies in early sensory areas which mirror not just response properties but also connectivity between pyramidal cells.

Quantitative models of auditory cortical processing

To generate insight from experimental data, it is critical to understand the inter-relationships between individual data points and place them in context within a structured framework. Quantitative modeling can provide the scaffolding for such an endeavor. Our main objective in this review is to provide a primer on the range of quantitative tools available to experimental auditory neuroscientists. Quantitative modeling is advantageous because it can provide a compact summary of observed data, make underlying assumptions explicit, and generate predictions for future experiments. Quantitative models may be developed to characterize or fit observed data, to test theories of how a task may be solved by neural circuits, to determine how observed biophysical details might contribute to measured activity patterns, or to predict how an experimental manipulation would affect neural activity. In complexity, quantitative models can range from those that are highly biophysically realistic and that include detailed simulations at the level of individual synapses, to those that use abstract and simplified neuron models to simulate entire networks. Here, we survey the landscape of recently developed models of auditory cortical processing, highlighting a small selection of models to demonstrate how they help generate insight into the mechanisms of auditory processing. We discuss examples ranging from models that use details of synaptic properties to explain the temporal pattern of cortical responses to those that use modern deep neural networks to gain insight into human fMRI data. We conclude by discussing a biologically realistic and interpretable model that our laboratory has developed to explore aspects of vocalization categorization in the auditory pathway.

Transcranial electrical stimulation: How can a simple conductor orchestrate complex brain activity?

Transcranial electrical stimulation (tES) is one of the oldest and yet least understood forms of brain stimulation. The idea that a weak electrical stimulus, applied outside the head, can meaningfully affect neural activity is often regarded as mysterious. Here, we argue that the direct effects of tES are not so mysterious: Extensive data from a wide range of model systems shows it has appreciable effects on the activity of individual neurons. Instead, the real mysteries are how tES interacts with the brain's own activity and how these dynamics can be controlled to produce desirable therapeutic effects. These are challenging problems, akin to repairing a complex machine while it is running, but they are not unique to tES or even neuroscience. We suggest that models of coupled oscillators, a common tool for studying interactions in other fields, may provide valuable insights. By combining these tools with our growing, interdisciplinary knowledge of brain dynamics, we are now in a good position to make progress in this area and meet the high demand for effective neuromodulation in neuroscience and psychiatry.

Distinct nonlinear spectrotemporal integration in primary and secondary auditory cortices

Animals sense sounds through hierarchical neural pathways that ultimately reach higher-order cortices to extract complex acoustic features, such as vocalizations. Elucidating how spectrotemporal integration varies along the hierarchy from primary to higher-order auditory cortices is a crucial step in understanding this elaborate sensory computation. Here we used two-photon calcium imaging and two-tone stimuli with various frequency-timing combinations to compare spectrotemporal integration between primary (A1) and secondary (A2) auditory cortices in mice. Individual neurons showed mixed supralinear and sublinear integration in a frequency-timing combination-specific manner, and we found unique integration patterns in these two areas. Temporally asymmetric spectrotemporal integration in A1 neurons enabled their discrimination of frequency-modulated sweep directions. In contrast, temporally symmetric and coincidence-preferring integration in A2 neurons made them ideal spectral integrators of concurrent multifrequency sounds. Moreover, the ensemble neural activity in A2 was sensitive to two-tone timings, and coincident two-tones evoked distinct ensemble activity patterns from the linear sum of component tones. Together, these results demonstrate distinct roles of A1 and A2 in encoding complex acoustic features, potentially suggesting parallel rather than sequential information extraction between these regions.

Creation of Neuronal Ensembles and Cell-Specific Homeostatic Plasticity through Chronic Sparse Optogenetic Stimulation

Cortical computations emerge from the dynamics of neurons embedded in complex cortical circuits. Within these circuits, neuronal ensembles, which represent subnetworks with shared functional connectivity, emerge in an experience-dependent manner. Here we induced ensembles in ex vivo cortical circuits from mice of either sex by differentially activating subpopulations through chronic optogenetic stimulation. We observed a decrease in voltage correlation, and importantly a synaptic decoupling between the stimulated and nonstimulated populations. We also observed a decrease in firing rate during Up-states in the stimulated population. These ensemble-specific changes were accompanied by decreases in intrinsic excitability in the stimulated population, and a decrease in connectivity between stimulated and nonstimulated pyramidal neurons. By incorporating the empirically observed changes in intrinsic excitability and connectivity into a spiking neural network model, we were able to demonstrate that changes in both intrinsic excitability and connectivity accounted for the decreased firing rate, but only changes in connectivity accounted for the observed decorrelation. Our findings help ascertain the mechanisms underlying the ability of chronic patterned stimulation to create ensembles within cortical circuits and, importantly, show that while Up-states are a global network-wide phenomenon, functionally distinct ensembles can preserve their identity during Up-states through differential firing rates and correlations.SIGNIFICANCE STATEMENT The connectivity and activity patterns of local cortical circuits are shaped by experience. This experience-dependent reorganization of cortical circuits is driven by complex interactions between different local learning rules, external input, and reciprocal feedback between many distinct brain areas. Here we used an ex vivo approach to demonstrate how simple forms of chronic external stimulation can shape local cortical circuits in terms of their correlated activity and functional connectivity. The absence of feedback between different brain areas and full control of external input allowed for a tractable system to study the underlying mechanisms and development of a computational model. Results show that differential stimulation of subpopulations of neurons significantly reshapes cortical circuits and forms subnetworks referred to as neuronal ensembles.

Mesoscopic description of hippocampal replay and metastability in spiking neural networks with short-term plasticity

Bottom-up models of functionally relevant patterns of neural activity provide an explicit link between neuronal dynamics and computation. A prime example of functional activity patterns are propagating bursts of place-cell activities called hippocampal replay, which is critical for memory consolidation. The sudden and repeated occurrences of these burst states during ongoing neural activity suggest metastable neural circuit dynamics. As metastability has been attributed to noise and/or slow fatigue mechanisms, we propose a concise mesoscopic model which accounts for both. Crucially, our model is bottom-up: it is analytically derived from the dynamics of finite-size networks of Linear-Nonlinear Poisson neurons with short-term synaptic depression. As such, noise is explicitly linked to stochastic spiking and network size, and fatigue is explicitly linked to synaptic dynamics. To derive the mesoscopic model, we first consider a homogeneous spiking neural network and follow the temporal coarse-graining approach of Gillespie to obtain a "chemical Langevin equation", which can be naturally interpreted as a stochastic neural mass model. The Langevin equation is computationally inexpensive to simulate and enables a thorough study of metastable dynamics in classical setups (population spikes and Up-Down-states dynamics) by means of phase-plane analysis. An extension of the Langevin equation for small network sizes is also presented. The stochastic neural mass model constitutes the basic component of our mesoscopic model for replay. We show that the mesoscopic model faithfully captures the statistical structure of individual replayed trajectories in microscopic simulations and in previously reported experimental data. Moreover, compared to the deterministic Romani-Tsodyks model of place-cell dynamics, it exhibits a higher level of variability regarding order, direction and timing of replayed trajectories, which seems biologically more plausible and could be functionally desirable. This variability is the product of a new dynamical regime where metastability emerges from a complex interplay between finite-size fluctuations and local fatigue.

Strong tuning for stereoscopic depth indicates orientation-specific recurrent circuitry in tree shrew V1

Neurons in the primary visual cortex (V1) are tuned to specific disparities between the two retinal images, which form the neural substrate for stereoscopic vision. We show that V1 neurons in tree shrews, but not in mice, display highly selective responses to narrow ranges of disparity in random-dot stereograms. Surprisingly, V1 neurons in both species show similarly strong tuning to gratings of varying interocular phase differences. This stimulus-dependent dissociation of disparity tuning can be explained by a network model that combines both feedforward and recurrent connections. The features of the model connections are supported by cortical organizations specific to each species. We validate this model by identifying putative inhibitory neurons and confirming their predicted disparity tuning in both species. Together, our studies establish a foundation for using tree shrews in studying binocular vision and raise an exciting possibility of how cortical columns could be uniquely important in computing stereoscopic depth.

Information maximization explains state-dependent synaptic plasticity and memory reorganization during non-rapid eye movement sleep

Slow waves during the non-rapid eye movement (NREM) sleep reflect the alternating up and down states of cortical neurons; global and local slow waves promote memory consolidation and forgetting, respectively. Furthermore, distinct spike-timing-dependent plasticity (STDP) operates in these up and down states. The contribution of different plasticity rules to neural information coding and memory reorganization remains unknown. Here, we show that optimal synaptic plasticity for information maximization in a cortical neuron model provides a unified explanation for these phenomena. The model indicates that the optimal synaptic plasticity is biased toward depression as the baseline firing rate increases. This property explains the distinct STDP observed in the up and down states. Furthermore, it explains how global and local slow waves predominantly potentiate and depress synapses, respectively, if the background firing rate of excitatory neurons declines with the spatial scale of waves as the model predicts. The model provides a unifying account of the role of NREM sleep, bridging neural information coding, synaptic plasticity, and memory reorganization.

Input correlations impede suppression of chaos and learning in balanced firing-rate networks

Neural circuits exhibit complex activity patterns, both spontaneously and evoked by external stimuli. Information encoding and learning in neural circuits depend on how well time-varying stimuli can control spontaneous network activity. We show that in firing-rate networks in the balanced state, external control of recurrent dynamics, i.e., the suppression of internally-generated chaotic variability, strongly depends on correlations in the input. A distinctive feature of balanced networks is that, because common external input is dynamically canceled by recurrent feedback, it is far more difficult to suppress chaos with common input into each neuron than through independent input. To study this phenomenon, we develop a non-stationary dynamic mean-field theory for driven networks. The theory explains how the activity statistics and the largest Lyapunov exponent depend on the frequency and amplitude of the input, recurrent coupling strength, and network size, for both common and independent input. We further show that uncorrelated inputs facilitate learning in balanced networks.

Object Boundary Detection in Natural Images May Depend on "Incitatory" Cell-Cell Interactions

Detecting object boundaries is crucial for recognition, but how the process unfolds in visual cortex remains unknown. To study the problem faced by a hypothetical boundary cell, and to predict how cortical circuitry could produce a boundary cell from a population of conventional "simple cells," we labeled 30,000 natural image patches and used Bayes' rule to help determine how a simple cell should influence a nearby boundary cell depending on its relative offset in receptive field position and orientation. We identified the following three basic types of cell-cell interactions: rising and falling interactions with a range of slopes and saturation rates, and nonmonotonic (bump-shaped) interactions with varying modes and amplitudes. Using simple models, we show that a ubiquitous cortical circuit motif consisting of direct excitation and indirect inhibition-a compound effect we call "incitation"-can produce the entire spectrum of simple cell-boundary cell interactions found in our dataset. Moreover, we show that the synaptic weights that parameterize an incitation circuit can be learned by a single-layer "delta" rule. We conclude that incitatory interconnections are a generally useful computing mechanism that the cortex may exploit to help solve difficult natural classification problems.SIGNIFICANCE STATEMENT Simple cells in primary visual cortex (V1) respond to oriented edges and have long been supposed to detect object boundaries, yet the prevailing model of a simple cell-a divisively normalized linear filter-is a surprisingly poor natural boundary detector. To understand why, we analyzed image statistics on and off object boundaries, allowing us to characterize the neural-style computations needed to perform well at this difficult natural classification task. We show that a simple circuit motif known to exist in V1 is capable of extracting high-quality boundary probability signals from local populations of simple cells. Our findings suggest a new, more general way of conceptualizing cell-cell interconnections in the cortex.

Key role of neuronal diversity in structured reservoir computing

Chaotic time series have been captured by reservoir computing models composed of a recurrent neural network whose output weights are trained in a supervised manner. These models, however, are typically limited to randomly connected networks of homogeneous units. Here, we propose a new class of structured reservoir models that incorporates a diversity of cell types and their known connections. In a first version of the model, the reservoir was composed of mean-rate units separated into pyramidal, parvalbumin, and somatostatin cells. Stability analysis of this model revealed two distinct dynamical regimes, namely, (i) an inhibition-stabilized network (ISN) where strong recurrent excitation is balanced by strong inhibition and (ii) a non-ISN network with weak excitation. These results were extended to a leaky integrate-and-fire model that captured different cell types along with their network architecture. ISN and non-ISN reservoir networks were trained to relay and generate a chaotic Lorenz attractor. Despite their increased performance, ISN networks operate in a regime of activity near the limits of stability where external perturbations yield a rapid divergence in output. The proposed framework of structured reservoir computing opens avenues for exploring how neural microcircuits can balance performance and stability when representing time series through distinct dynamical regimes.