The geometry of representational drift in natural and artificial neural networks.

Neurons in sensory areas encode/represent stimuli. Surprisingly, recent studies have suggested that, even during persistent performance, these representations are not stable and change over the course of days and weeks. We examine stimulus representations from fluorescence recordings across hundreds of neurons in the visual cortex using in vivo two-photon calcium imaging and we corroborate previous studies finding that such representations change as experimental trials are repeated across days. This phenomenon has been termed "representational drift". In this study we geometrically characterize the properties of representational drift in the primary visual cortex of mice in two open datasets from the Allen Institute and propose a potential mechanism behind such drift. We observe representational drift both for passively presented stimuli, as well as for stimuli which are behaviorally relevant. Across experiments, the drift differs from in-session variance and most often occurs along directions that have the most in-class variance, leading to a significant turnover in the neurons used for a given representation. Interestingly, despite this significant change due to drift, linear classifiers trained to distinguish neuronal representations show little to no degradation in performance across days. The features we observe in the neural data are similar to properties of artificial neural networks where representations are updated by continual learning in the presence of dropout, i.e. a random masking of nodes/weights, but not other types of noise. Therefore, we conclude that a potential reason for the representational drift in biological networks is driven by an underlying dropout-like noise while continuously learning and that such a mechanism may be computational advantageous for the brain in the same way it is for artificial neural networks, e.g. preventing overfitting.

Adaptation supports short-term memory in a visual change detection task.

The maintenance of short-term memories is critical for survival in a dynamically changing world. Previous studies suggest that this memory can be stored in the form of persistent neural activity or using a synaptic mechanism, such as with short-term plasticity. Here, we compare the predictions of these two mechanisms to neural and behavioral measurements in a visual change detection task. Mice were trained to respond to changes in a repeated sequence of natural images while neural activity was recorded using two-photon calcium imaging. We also trained two types of artificial neural networks on the same change detection task as the mice. Following fixed pre-processing using a pretrained convolutional neural network, either a recurrent neural network (RNN) or a feedforward neural network with short-term synaptic depression (STPNet) was trained to the same level of performance as the mice. While both networks are able to learn the task, the STPNet model contains units whose activity are more similar to the in vivo data and produces errors which are more similar to the mice. When images are omitted, an unexpected perturbation which was absent during training, mice often do not respond to the omission but are more likely to respond to the subsequent image. Unlike the RNN model, STPNet produces a similar pattern of behavior. These results suggest that simple neural adaptation mechanisms may serve as an important bottom-up memory signal in this task, which can be used by downstream areas in the decision-making process.

High-density extracellular probes reveal dendritic backpropagation and facilitate neuron classification.

Different neuron types serve distinct roles in neural processing. Extracellular electrical recordings are extensively used to study brain function but are typically blind to cell identity. Morphoelectrical properties of neurons measured on spatially dense electrode arrays have the potential to distinguish neuron types. We used high-density silicon probes to record from cortical and subcortical regions of the mouse brain. Extracellular waveforms of each neuron were detected across many channels and showed distinct spatiotemporal profiles among brain regions. Classification of neurons by brain region was improved with multichannel compared with single-channel waveforms. In visual cortex, unsupervised clustering identified the canonical regular-spiking (RS) and fast-spiking (FS) classes but also indicated a subclass of RS units with unidirectional backpropagating action potentials (BAPs). Moreover, BAPs were observed in many hippocampal RS cells. Overall, waveform analysis of spikes from high-density probes aids neuron identification and can reveal dendritic backpropagation. NEW & NOTEWORTHY It is challenging to identify neuron types with extracellular electrophysiology in vivo. We show that spatiotemporal action potentials measured on high-density electrode arrays can capture cell type-specific morphoelectrical properties, allowing classification of neurons across brain structures and within the cortex. Moreover, backpropagating action potentials are reliably detected in vivo from subpopulations of cortical and hippocampal neurons. Together, these results enhance the utility of dense extracellular electrophysiology for cell-type interrogation of brain network function.

Visual physiology of the layer 4 cortical circuit in silico.

Despite advances in experimental techniques and accumulation of large datasets concerning the composition and properties of the cortex, quantitative modeling of cortical circuits under in-vivo-like conditions remains challenging. Here we report and publicly release a biophysically detailed circuit model of layer 4 in the mouse primary visual cortex, receiving thalamo-cortical visual inputs. The 45,000-neuron model was subjected to a battery of visual stimuli, and results were compared to published work and new in vivo experiments. Simulations reproduced a variety of observations, including effects of optogenetic perturbations. Critical to the agreement between responses in silico and in vivo were the rules of functional synaptic connectivity between neurons. Interestingly, after extreme simplification the model still performed satisfactorily on many measurements, although quantitative agreement with experiments suffered. These results emphasize the importance of functional rules of cortical wiring and enable a next generation of data-driven models of in vivo neural activity and computations.

Inferring cortical function in the mouse visual system through large-scale systems neuroscience.

The scientific mission of the Project MindScope is to understand neocortex, the part of the mammalian brain that gives rise to perception, memory, intelligence, and consciousness. We seek to quantitatively evaluate the hypothesis that neocortex is a relatively homogeneous tissue, with smaller functional modules that perform a common computational function replicated across regions. We here focus on the mouse as a mammalian model organism with genetics, physiology, and behavior that can be readily studied and manipulated in the laboratory. We seek to describe the operation of cortical circuitry at the computational level by comprehensively cataloging and characterizing its cellular building blocks along with their dynamics and their cell type-specific connectivities. The project is also building large-scale experimental platforms (i.e., brain observatories) to record the activity of large populations of cortical neurons in behaving mice subject to visual stimuli. A primary goal is to understand the series of operations from visual input in the retina to behavior by observing and modeling the physical transformations of signals in the corticothalamic system. We here focus on the contribution that computer modeling and theory make to this long-term effort.

New Breakthroughs in Understanding the Role of Functional Interactions between the Neocortex and the Claustrum.

Almost all areas of the neocortex are connected with the claustrum, a nucleus located between the neocortex and the striatum, yet the functions of corticoclaustral and claustrocortical connections remain largely obscure. As major efforts to model the neocortex are currently underway, it has become increasingly important to incorporate the corticoclaustral system into theories of cortical function. This Mini-Symposium was motivated by a series of recent studies which have sparked new hypotheses regarding the function of claustral circuits. Anatomical, ultrastructural, and functional studies indicate that the claustrum is most highly interconnected with prefrontal cortex, suggesting important roles in higher cognitive processing, and that the organization of the corticoclaustral system is distinct from the driver/modulator framework often used to describe the corticothalamic system. Recent findings supporting roles in detecting novel sensory stimuli, directing attention and setting behavioral states, were the subject of the Mini-Symposium at the 2017 Society for Neuroscience Annual Meeting.

Gain control by layer six in cortical circuits of vision.

After entering the cerebral cortex, sensory information spreads through six different horizontal neuronal layers that are interconnected by vertical axonal projections. It is believed that through these projections layers can influence each other's response to sensory stimuli, but the specific role that each layer has in cortical processing is still poorly understood. Here we show that layer six in the primary visual cortex of the mouse has a crucial role in controlling the gain of visually evoked activity in neurons of the upper layers without changing their tuning to orientation. This gain modulation results from the coordinated action of layer six intracortical projections to superficial layers and deep projections to the thalamus, with a substantial role of the intracortical circuit. This study establishes layer six as a major mediator of cortical gain modulation and suggests that it could be a node through which convergent inputs from several brain areas can regulate the earliest steps of cortical visual processing.

Lateral presynaptic inhibition mediates gain control in an olfactory circuit.

Olfactory signals are transduced by a large family of odorant receptor proteins, each of which corresponds to a unique glomerulus in the first olfactory relay of the brain. Crosstalk between glomeruli has been proposed to be important in olfactory processing, but it is not clear how these interactions shape the odour responses of second-order neurons. In the Drosophila antennal lobe (a region analogous to the vertebrate olfactory bulb), we selectively removed most interglomerular input to genetically identified second-order olfactory neurons. Here we show that this broadens the odour tuning of these neurons, implying that interglomerular inhibition dominates over interglomerular excitation. The strength of this inhibitory signal scales with total feedforward input to the entire antennal lobe, and has similar tuning in different glomeruli. A substantial portion of this interglomerular inhibition acts at a presynaptic locus, and our results imply that this is mediated by both ionotropic and metabotropic receptors on the same nerve terminal.

Simple synaptic modulations implement diverse novelty computations.

Detecting novelty is ethologically useful for an organism's survival. Recent experiments characterize how different types of novelty over timescales from seconds to weeks are reflected in the activity of excitatory and inhibitory neuron types. Here, we introduce a learning mechanism, familiarity-modulated synapses (FMSs), consisting of multiplicative modulations dependent on presynaptic or pre/postsynaptic neuron activity. With FMSs, network responses that encode novelty emerge under unsupervised continual learning and minimal connectivity constraints. Implementing FMSs within an experimentally constrained model of a visual cortical circuit, we demonstrate the generalizability of FMSs by simultaneously fitting absolute, contextual, and omission novelty effects. Our model also reproduces functional diversity within cell subpopulations, leading to experimentally testable predictions about connectivity and synaptic dynamics that can produce both population-level novelty responses and heterogeneous individual neuron signals. Altogether, our findings demonstrate how simple plasticity mechanisms within a cortical circuit structure can produce qualitatively distinct and complex novelty responses.

Backward masking in mice requires visual cortex.

Visual masking can reveal the timescale of perception, but the underlying circuit mechanisms are not understood. Here we describe a backward masking task in mice and humans in which the location of a stimulus is potently masked. Humans report reduced subjective visibility that tracks behavioral deficits. In mice, both masking and optogenetic silencing of visual cortex (V1) reduce performance over a similar timecourse but have distinct effects on response rates and accuracy. Activity in V1 is consistent with masked behavior when quantified over long, but not short, time windows. A dual accumulator model recapitulates both mouse and human behavior. The model and subjects' performance imply that the initial spikes in V1 can trigger a correct response, but subsequent V1 activity degrades performance. Supporting this hypothesis, optogenetically suppressing mask-evoked activity in V1 fully restores accurate behavior. Together, these results demonstrate that mice, like humans, are susceptible to masking and that target and mask information is first confounded downstream of V1.

Behavioral strategy shapes activation of the Vip-Sst disinhibitory circuit in visual cortex.

In complex environments, animals can adopt diverse strategies to find rewards. How distinct strategies differentially engage brain circuits is not well understood. Here, we investigate this question, focusing on the cortical Vip-Sst disinhibitory circuit between vasoactive intestinal peptide-postive (Vip) interneurons and somatostatin-positive (Sst) interneurons. We characterize the behavioral strategies used by mice during a visual change detection task. Using a dynamic logistic regression model, we find that individual mice use mixtures of a visual comparison strategy and a statistical timing strategy. Separately, mice also have periods of task engagement and disengagement. Two-photon calcium imaging shows large strategy-dependent differences in neural activity in excitatory, Sst inhibitory, and Vip inhibitory cells in response to both image changes and image omissions. In contrast, task engagement has limited effects on neural population activity. We find that the diversity of neural correlates of strategy can be understood parsimoniously as the increased activation of the Vip-Sst disinhibitory circuit during the visual comparison strategy, which facilitates task-appropriate responses.

Ultra-high density electrodes improve detection, yield, and cell type identification in neuronal recordings.

To understand the neural basis of behavior, it is essential to sensitively and accurately measure neural activity at single neuron and single spike resolution. Extracellular electrophysiology delivers this, but it has biases in the neurons it detects and it imperfectly resolves their action potentials. To minimize these limitations, we developed a silicon probe with much smaller and denser recording sites than previous designs, called Neuropixels Ultra (NP Ultra). This device samples neuronal activity at ultra-high spatial density (~10 times higher than previous probes) with low noise levels, while trading off recording span. NP Ultra is effectively an implantable voltage-sensing camera that captures a planar image of a neuron's electrical field. We use a spike sorting algorithm optimized for these probes to demonstrate that the yield of visually-responsive neurons in recordings from mouse visual cortex improves up to ~3-fold. We show that NP Ultra can record from small neuronal structures including axons and dendrites. Recordings across multiple brain regions and four species revealed a subset of extracellular action potentials with unexpectedly small spatial spread and axon-like features. We share a large-scale dataset of these brain-wide recordings in mice as a resource for studies of neuronal biophysics. Finally, using ground-truth identification of three major inhibitory cortical cell types, we found that these cell types were discriminable with approximately 75% success, a significant improvement over lower-resolution recordings. NP Ultra improves spike sorting performance, detection of subcellular compartments, and cell type classification to enable more powerful dissection of neural circuit activity during behavior.

SHIELD: Skull-shaped hemispheric implants enabling large-scale electrophysiology datasets in the mouse brain.

To understand the neural basis of behavior, it is essential to measure spiking dynamics across many interacting brain regions. Although new technologies, such as Neuropixels probes, facilitate multi-regional recordings, significant surgical and procedural hurdles remain for these experiments to achieve their full potential. Here, we describe skull-shaped hemispheric implants enabling large-scale electrophysiology datasets (SHIELD). These 3D-printed skull-replacement implants feature customizable insertion holes, allowing dozens of cortical and subcortical structures to be recorded in a single mouse using repeated multi-probe insertions over many days. We demonstrate the procedure's high success rate, biocompatibility, lack of adverse effects on behavior, and compatibility with imaging and optogenetics. To showcase SHIELD's scientific utility, we use multi-probe recordings to reveal novel insights into how alpha rhythms organize spiking activity across visual and sensorimotor networks. Overall, this method enables powerful, large-scale electrophysiological experiments for the study of distributed neural computation.

Familiarity modulated synapses model visual cortical circuit novelty responses.

Since environments are constantly in flux, the brain's ability to identify novel stimuli that fall outside its own internal representation of the world is crucial for an organism's survival. Within the mammalian neocortex, inhibitory microcircuits are proposed to regulate activity in an experience-dependent manner and different inhibitory neuron subtypes exhibit distinct novelty responses. Discerning the function of diverse neural circuits and their modulation by experience can be daunting unless one has a biologically plausible mechanism to detect and learn from novel experiences that is both understandable and flexible. Here we introduce a learning mechanism, familiarity modulated synapses (FMSs), through which a network response that encodes novelty emerges from unsupervised synaptic modifications depending only on the presynaptic or both the pre- and postsynaptic activity. FMSs stand apart from other familiarity mechanisms in their simplicity: they operate under continual learning, do not require specialized architecture, and can distinguish novelty rapidly without requiring feedback. Implementing FMSs within a model of a visual cortical circuit that includes multiple inhibitory populations, we simultaneously reproduce three distinct novelty effects recently observed in experimental data from visual cortical circuits in mice: absolute, contextual, and omission novelty. Additionally, our model results in a set of diverse physiological responses across cell subpopulations, allowing us to analyze how their connectivity and synaptic dynamics influences their distinct behavior, leading to predictions that can be tested in experiment. Altogether, our findings demonstrate how experimentally-constrained cortical circuit structure can give rise to qualitatively distinct novelty responses using simple plasticity mechanisms. The flexibility of FMSs opens the door to computationally and theoretically investigating how distinct synapse modulations can lead to a variety of experience-dependent responses in a simple, understandable, and biologically plausible setup.

Influence of claustrum on cortex varies by area, layer, and cell type.

The claustrum is a small subcortical structure with widespread connections to disparate regions of the cortex. However, the impact of the claustrum on cortical activity is not fully understood, particularly beyond frontal areas. Here, using optogenetics and multi-regional Neuropixels recordings from over 15,000 cortical neurons in awake mice, we demonstrate that the effect of claustrum input to the cortex differs depending on brain area, layer, and cell type. Brief claustrum stimulation, producing approximately 1 spike per claustrum neuron, affects many fast spiking (FS; putative inhibitory) but relatively fewer regular-spiking (RS; putative excitatory) cortical neurons and leads to a modest decrease in population activity in frontal cortical areas. Prolonged claustrum stimulation affects many more cortical neurons and can increase or decrease spiking activity. More excitation occurs in posterior regions and superficial layers, while inhibition predominates in frontal regions and deeper layers. These findings suggest that claustro-cortical circuits are organized into functional modules.

Coordinated changes in a cortical circuit sculpt effects of novelty on neural dynamics.

Recent studies have found dramatic cell-type specific responses to stimulus novelty, highlighting the importance of analyzing the cortical circuitry at the cell-type specific level of granularity to understand brain function. Although initial work classified and characterized activity for each cell type, the specific alterations in cortical circuitry-particularly when multiple novelty effects interact-remain unclear. To address this gap, we employed a large-scale public dataset of electrophysiological recordings in the visual cortex of awake, behaving mice using Neuropixels probes and designed population network models to investigate the observed changes in neural dynamics in response to a combination of distinct forms of novelty. The model parameters were rigorously constrained by publicly available structural datasets, including multi-patch synaptic physiology and electron microscopy data. Our systematic optimization approach identified tens of thousands of model parameter sets that replicate the observed neural activity. Analysis of these solutions revealed generally weaker connections under novel stimuli, as well as a shift in the balance e between SST and VIP populations. Along with this, PV and SST populations experienced overall more excitatory influences compared to excitatory and VIP populations. Our results also highlight the role of VIP neurons in multiple aspects of visual stimulus processing and altering gain and saturation dynamics under novel conditions. In sum, our findings provide a systematic characterization of how the cortical circuit adapts to stimulus novelty by combining multiple rich public datasets.

Cortico-thalamo-cortical interactions modulate electrically evoked EEG responses in mice.

Perturbational complexity analysis predicts the presence of consciousness in volunteers and patients by stimulating the brain with brief pulses, recording EEG responses, and computing their spatiotemporal complexity. We examined the underlying neural circuits in mice by directly stimulating cortex while recording with EEG and Neuropixels probes during wakefulness and isoflurane anesthesia. When mice are awake, stimulation of deep cortical layers reliably evokes locally a brief pulse of excitation, followed by a biphasic sequence of 120 ms profound off period and a rebound excitation. A similar pattern, partially attributed to burst spiking, is seen in thalamic nuclei and is associated with a pronounced late component in the evoked EEG. We infer that cortico-thalamo-cortical interactions drive the long-lasting evoked EEG signals elicited by deep cortical stimulation during the awake state. The cortical and thalamic off period and rebound excitation, and the late component in the EEG, are reduced during running and absent during anesthesia.

Uncovering circuit mechanisms of current sinks and sources with biophysical simulations of primary visual cortex.

Local field potential (LFP) recordings reflect the dynamics of the current source density (CSD) in brain tissue. The synaptic, cellular, and circuit contributions to current sinks and sources are ill-understood. We investigated these in mouse primary visual cortex using public Neuropixels recordings and a detailed circuit model based on simulating the Hodgkin-Huxley dynamics of >50,000 neurons belonging to 17 cell types. The model simultaneously captured spiking and CSD responses and demonstrated a two-way dissociation: firing rates are altered with minor effects on the CSD pattern by adjusting synaptic weights, and CSD is altered with minor effects on firing rates by adjusting synaptic placement on the dendrites. We describe how thalamocortical inputs and recurrent connections sculpt specific sinks and sources early in the visual response, whereas cortical feedback crucially alters them in later stages. These results establish quantitative links between macroscopic brain measurements (LFP/CSD) and microscopic biophysics-based understanding of neuron dynamics and show that CSD analysis provides powerful constraints for modeling beyond those from considering spikes.

Acute head-fixed recordings in awake mice with multiple Neuropixels probes.

Multi-electrode arrays such as Neuropixels probes enable electrophysiological recordings from large populations of single neurons with high temporal resolution. By using such probes, the activity from functionally interacting, yet distinct, brain regions can be measured simultaneously by inserting multiple probes into the same subject. However, the use of multiple probes in small animals such as mice requires the removal of a sizable fraction of the skull, while also minimizing tissue damage and keeping the brain stable during the recordings. Here, we describe a step-by-step process designed to facilitate reliable recordings from up to six Neuropixels probes simultaneously in awake, head-fixed mice. The procedure involves four stages: the implantation of a headframe and a removable glass coverslip, the precise positioning of the Neuropixels probes at targeted points on the brain surface, the placement of a perforated plastic imaging window and the insertion of the probes into the brain of an awake mouse. The approach provides access to multiple brain regions and has been successfully applied across hundreds of mice. The procedure has been optimized for dense recordings from the mouse visual system, but it can be adapted for alternative recording configurations to target multiple probes in other brain areas. The protocol is suitable for users with experience in stereotaxic surgery in mice.

A survey of neurophysiological differentiation across mouse visual brain areas and timescales.

Neurophysiological differentiation (ND), a measure of the number of distinct activity states that a neural population visits over a time interval, has been used as a correlate of meaningfulness or subjective perception of visual stimuli. ND has largely been studied in non-invasive human whole-brain recordings where spatial resolution is limited. However, it is likely that perception is supported by discrete neuronal populations rather than the whole brain. Therefore, here we use Neuropixels recordings from the mouse brain to characterize the ND metric across a wide range of temporal scales, within neural populations recorded at single-cell resolution in localized regions. Using the spiking activity of thousands of simultaneously recorded neurons spanning 6 visual cortical areas and the visual thalamus, we show that the ND of stimulus-evoked activity of the entire visual cortex is higher for naturalistic stimuli relative to artificial ones. This finding holds in most individual areas throughout the visual hierarchy. Moreover, for animals performing an image change detection task, ND of the entire visual cortex (though not individual areas) is higher for successful detection compared to failed trials, consistent with the assumed perception of the stimulus. Together, these results suggest that ND computed on cellular-level neural recordings is a useful tool highlighting cell populations that may be involved in subjective perception.