Learning Enhances Sensory and Multiple Non-sensory Representations in Primary Visual Cortex

A neural circuit architecture for rapid learning in goal-directed navigation

Anchoring goals to spatial representations enables flexible navigation but is challenging in novel environments when both representations must be acquired simultaneously. We propose a framework for how Drosophila uses internal representations of head direction (HD) to build goal representations upon selective thermal reinforcement. We show that flies use stochastically generated fixations and directed saccades to express heading preferences in an operant visual learning paradigm and that HD neurons are required to modify these preferences based on reinforcement. We used a symmetric visual setting to expose how flies' HD and goal representations co-evolve and how the reliability of these interacting representations impacts behavior. Finally, we describe how rapid learning of new goal headings may rest on a behavioral policy whose parameters are flexible but whose form is genetically encoded in circuit architecture. Such evolutionarily structured architectures, which enable rapidly adaptive behavior driven by internal representations, may be relevant across species.

Goal-directed learning is multidimensional and accompanied by diverse and widespread changes in neocortical signaling

New tasks are often learned in stages with each stage reflecting a different learning challenge. Accordingly, each learning stage is likely mediated by distinct neuronal processes. And yet, most rodent studies of the neuronal correlates of goal-directed learning focus on individual outcome measures and individual brain regions. Here, we longitudinally studied mice from naïve to expert performance in a head-fixed, operant conditioning whisker discrimination task. In addition to tracking the primary behavioral outcome of stimulus discrimination, we tracked and compared an array of object-based and temporal-based behavioral measures. These behavioral analyses identify multiple, partially overlapping learning stages in this task, consistent with initial response implementation, early stimulus-response generalization, and late response inhibition. To begin to understand the neuronal foundations of these learning processes, we performed widefield Ca2+ imaging of dorsal neocortex throughout learning and correlated behavioral measures with neuronal activity. We found distinct and widespread correlations between neocortical activation patterns and various behavioral measures. For example, improvements in sensory discrimination correlated with target stimulus evoked activations of response-related cortices along with distractor stimulus evoked global cortical suppression. Our study reveals multidimensional learning for a simple goal-directed learning task and generates hypotheses for the neuronal modulations underlying these various learning processes.

Distinctive Neurophysiological Signatures of Analgesia after Inflammatory Pain in the ACC of Freely Moving Mice

Preclinical assessments of pain have often relied upon behavioral measurements and anesthetized neurophysiological recordings. Current technologies enabling large-scale neural recordings, however, have the potential to unveil quantifiable pain signals in conscious animals for preclinical studies. Although pain processing is distributed across many brain regions, the anterior cingulate cortex (ACC) is of particular interest in isolating these signals given its suggested role in the affective ("unpleasant") component of pain. Here, we explored the utility of the ACC toward preclinical pain research using head-mounted miniaturized microscopes to record calcium transients in freely moving male mice expressing genetically encoded calcium indicator 6f (GCaMP6f) under the Thy1 promoter. We verified the expression of GCaMP6f in excitatory neurons and found no intrinsic behavioral differences in this model. Using a multimodal stimulation paradigm across naive, pain, and analgesic conditions, we found that while ACC population activity roughly scaled with stimulus intensity, single-cell representations were highly flexible. We found only low-magnitude increases in population activity after complete Freund's adjuvant (CFA) and insufficient evidence for the existence of a robust nociceptive ensemble in the ACC. However, we found a temporal sharpening of response durations and generalized increases in pairwise neural correlations in the presence of the mechanistically distinct analgesics gabapentin or ibuprofen after (but not before) CFA-induced inflammatory pain. This increase was not explainable by changes in locomotion alone. Taken together, these results highlight challenges in isolating distinct pain signals among flexible representations in the ACC but suggest a neurophysiological hallmark of analgesia after pain that generalizes to at least two analgesics.

Anterior cingulate cortex provides the neural substrates for feedback-driven iteration of decision and value representation

Adjusting decision-making under uncertain and dynamic situations is the hallmark of intelligence. It requires a system capable of converting feedback information to renew the internal value. The anterior cingulate cortex (ACC) involves in error and reward events that prompt switching or maintenance of current decision strategies. However, it is unclear whether and how the changes of stimulus-action mapping during behavioral adaptation are encoded, nor how such computation drives decision adaptation. Here, we tracked ACC activity in male mice performing go/no-go auditory discrimination tasks with manipulated stimulus-reward contingencies. Individual ACC neurons integrate the outcome information to the value representation in the next-run trials. Dynamic recruitment of them determines the learning rate of error-guided value iteration and decision adaptation, forming a non-linear feedback-driven updating system to secure the appropriate decision switch. Optogenetically suppressing ACC significantly slowed down feedback-driven decision switching without interfering with the execution of the established strategy.

A mathematical theory of relational generalization in transitive inference

Humans and animals routinely infer relations between different items or events and generalize these relations to novel combinations of items. This allows them to respond appropriately to radically novel circumstances and is fundamental to advanced cognition. However, how learning systems (including the brain) can implement the necessary inductive biases has been unclear. We investigated transitive inference (TI), a classic relational task paradigm in which subjects must learn a relation ([Formula: see text] and [Formula: see text]) and generalize it to new combinations of items ([Formula: see text]). Through mathematical analysis, we found that a broad range of biologically relevant learning models (e.g. gradient flow or ridge regression) perform TI successfully and recapitulate signature behavioral patterns long observed in living subjects. First, we found that models with item-wise additive representations automatically encode transitive relations. Second, for more general representations, a single scalar "conjunctivity factor" determines model behavior on TI and, further, the principle of norm minimization (a standard statistical inductive bias) enables models with fixed, partly conjunctive representations to generalize transitively. Finally, neural networks in the "rich regime," which enables representation learning and improves generalization on many tasks, unexpectedly show poor generalization and anomalous behavior on TI. We find that such networks implement a form of norm minimization (over hidden weights) that yields a local encoding mechanism lacking transitivity. Our findings show how minimal statistical learning principles give rise to a classical relational inductive bias (transitivity), explain empirically observed behaviors, and establish a formal approach to understanding the neural basis of relational abstraction.

Acquisition of non-olfactory encoding improves odour discrimination in olfactory cortex

Olfaction is influenced by contextual factors, past experiences, and the animal's internal state. Whether this information is integrated at the initial stages of cortical odour processing is not known, nor how these signals may influence odour encoding. Here we revealed multiple and diverse non-olfactory responses in the primary olfactory (piriform) cortex (PCx), which dynamically enhance PCx odour discrimination according to behavioural demands. We performed recordings of PCx neurons from mice trained in a virtual reality task to associate odours with visual contexts to obtain a reward. We found that learning shifts PCx activity from encoding solely odours to a regime in which positional, contextual, and associative responses emerge on odour-responsive neurons that become mixed-selective. The modulation of PCx activity by these non-olfactory signals was dynamic, improving odour decoding during task engagement and in rewarded contexts. This improvement relied on the acquired mixed-selectivity, demonstrating how integrating extra-sensory inputs in sensory cortices can enhance sensory processing while encoding the behavioural relevance of stimuli.

Stimulus information guides the emergence of behavior-related signals in primary somatosensory cortex during learning

Neurons in the primary cortex carry sensory- and behavior-related information, but it remains an open question how this information emerges and intersects together during learning. Current evidence points to two possible learning-related changes: sensory information increases in the primary cortex or sensory information remains stable, but its readout efficiency in association cortices increases. We investigated this question by imaging neuronal activity in mouse primary somatosensory cortex before, during, and after learning of an object localization task. We quantified sensory- and behavior-related information and estimated how much sensory information was used to instruct perceptual choices as learning progressed. We find that sensory information increases from the start of training, while choice information is mostly present in the later stages of learning. Additionally, the readout of sensory information becomes more efficient with learning as early as in the primary sensory cortex. Together, our results highlight the importance of primary cortical neurons in perceptual learning.

Rapid emergence of latent knowledge in the sensory cortex drives learning

Rapid learning confers significant advantages to animals in ecological environments. Despite the need for speed, animals appear to only slowly learn to associate rewarded actions with predictive cues1-4. This slow learning is thought to be supported by a gradual expansion of predictive cue representation in the sensory cortex2,5. However, evidence is growing that animals learn more rapidly than classical performance measures suggest6-8, challenging the prevailing model of sensory cortical plasticity. Here, we investigated the relationship between learning and sensory cortical representations. We trained mice on an auditory go/no-go task that dissociated the rapid acquisition of task contingencies (learning) from its slower expression (performance)7. Optogenetic silencing demonstrated that the auditory cortex (AC) drives both rapid learning and slower performance gains but becomes dispensable at expert. Rather than enhancement or expansion of cue representations9, two-photon calcium imaging of AC excitatory neurons throughout learning revealed two higher-order signals that were causal to learning and performance. First, a reward prediction (RP) signal emerged rapidly within tens of trials, was present after action-related errors only early in training, and faded at expert levels. Strikingly, silencing at the time of the RP signal impaired rapid learning, suggesting it serves an associative and teaching role. Second, a distinct cell ensemble encoded and controlled licking suppression that drove the slower performance improvements. These two ensembles were spatially clustered but uncoupled from underlying sensory representations, indicating a higher-order functional segregation within AC. Our results reveal that the sensory cortex manifests higher-order computations that separably drive rapid learning and slower performance improvements, reshaping our understanding of the fundamental role of the sensory cortex.

Dimensionality reduction beyond neural subspaces with slice tensor component analysis

Recent work has argued that large-scale neural recordings are often well described by patterns of coactivation across neurons. Yet the view that neural variability is constrained to a fixed, low-dimensional subspace may overlook higher-dimensional structure, including stereotyped neural sequences or slowly evolving latent spaces. Here we argue that task-relevant variability in neural data can also cofluctuate over trials or time, defining distinct 'covariability classes' that may co-occur within the same dataset. To demix these covariability classes, we develop sliceTCA (slice tensor component analysis), a new unsupervised dimensionality reduction method for neural data tensors. In three example datasets, including motor cortical activity during a classic reaching task in primates and recent multiregion recordings in mice, we show that sliceTCA can capture more task-relevant structure in neural data using fewer components than traditional methods. Overall, our theoretical framework extends the classic view of low-dimensional population activity by incorporating additional classes of latent variables capturing higher-dimensional structure.

Transient enhancement of stimulus-evoked activity in neocortex during sensory learning

Synaptic potentiation has been linked to learning in sensory cortex, but the connection between this potentiation and increased sensory-evoked neural activity is not clear. Here, we used longitudinal in vivo Ca2+ imaging in the barrel cortex of awake mice to test the hypothesis that increased excitatory synaptic strength during the learning of a whisker-dependent sensory-association task would be correlated with enhanced stimulus-evoked firing. To isolate stimulus-evoked responses from dynamic, task-related activity, imaging was performed outside of the training context. Although prior studies indicate that multiwhisker stimuli drive robust subthreshold activity, we observed sparse activation of L2/3 pyramidal (Pyr) neurons in both control and trained mice. Despite evidence for excitatory synaptic strengthening at thalamocortical and intracortical synapses in this brain area at the onset of learning-indeed, under our imaging conditions thalamocortical axons were robustly activated-we observed that L2/3 Pyr neurons in somatosensory (barrel) cortex displayed only modest increases in stimulus-evoked activity that were concentrated at the onset of training. Activity renormalized over longer training periods. In contrast, when stimuli and rewards were uncoupled in a pseudotraining paradigm, stimulus-evoked activity in L2/3 Pyr neurons was significantly suppressed. These findings indicate that sensory-association training but not sensory stimulation without coupled rewards may briefly enhance sensory-evoked activity, a phenomenon that might help link sensory input to behavioral outcomes at the onset of learning.

Reach-to-Grasp and tactile discrimination task: A new task for the study of sensory-motor learning

Although active touch in rodents arises from the forepaws as well as whiskers, most research on active touch only focuses on whiskers. This results in a paucity of tasks designed to assess the process of active touch with a forepaw. We develop a new experimental task, the Reach-to-Grasp and Tactile Discrimination task (RGTD task), to examine active touch with a forepaw in rodents, particularly changes in processes of active touch during motor skill learning. In the RGTD task, animals are required to (1) extend their forelimb to an object, (2) grasp the object, and (3) manipulate the grasped object with the forelimb. The animals must determine the direction of the manipulation based on active touch sensations arising during the period of the grasping. In experiment 1 of the present study, we showed that rats can learn the RGTD task. In experiment 2, we confirmed that the rats are capable of reversal learning of the RGTD task. The RGTD task shared most of the reaching movements involved with conventional forelimb reaching tasks. From the standpoint of a discrimination task, the RGTD task enables rigorous experimental control, for example by removing bias in the stimulus-response correspondence, and makes it possible to utilize diverse experimental procedures that have been difficult in prior tasks.

Differential representation of sensory information and behavioral choice across layers of the mouse auditory cortex

The activity of neurons in sensory areas sometimes covaries with upcoming choices in decision-making tasks. However, the prevalence, causal origin, and functional role of choice-related activity remain controversial. Understanding the circuit-logic of decision signals in sensory areas will require understanding their laminar specificity, but simultaneous recordings of neural activity across the cortical layers in forced-choice discrimination tasks have not yet been performed. Here, we describe neural activity from such recordings in the auditory cortex of mice during a frequency discrimination task with delayed report, which, as we show, requires the auditory cortex. Stimulus-related information was widely distributed across layers but disappeared very quickly after stimulus offset. Choice selectivity emerged toward the end of the delay period-suggesting a top-down origin-but only in the deep layers. Early stimulus-selective and late choice-selective deep neural ensembles were correlated, suggesting that the choice-selective signal fed back to the auditory cortex is not just action specific but develops as a consequence of the sensory-motor contingency imposed by the task.

Learning to learn: Single session acquisition of new rules by freely moving mice

Learning from examples and adapting to new circumstances are fundamental attributes of human cognition. However, it is unclear what conditions allow for fast and successful learning, especially in nonhuman subjects. To determine how rapidly freely moving mice can learn a new discrimination criterion (DC), we design a two-alternative forced-choice visual discrimination paradigm in which the DCs governing the task can change between sessions. We find that experienced animals can learn a new DC after being exposed to only five training and three testing trials. The propensity for single session learning improves over time and is accurately predicted based on animal experience and criterion difficulty. After establishing the procedural learning of a paradigm, mice continuously improve their performance in new circumstances. Thus, mice learn to learn.

Learning leaves a memory trace in motor cortex

How are we able to learn new behaviors without disrupting previously learned ones? To understand how the brain achieves this, we used a brain-computer interface (BCI) learning paradigm, which enables us to detect the presence of a memory of one behavior while performing another. We found that learning to use a new BCI map altered the neural activity that monkeys produced when they returned to using a familiar BCI map in a way that was specific to the learning experience. That is, learning left a "memory trace" in the primary motor cortex. This memory trace coexisted with proficient performance under the familiar map, primarily by altering neural activity in dimensions that did not impact behavior. Forming memory traces might be how the brain is able to provide for the joint learning of multiple behaviors without interference.

A mathematical theory of relational generalization in transitive inference

Humans and animals routinely infer relations between different items or events and generalize these relations to novel combinations of items. This allows them to respond appropriately to radically novel circumstances and is fundamental to advanced cognition. However, how learning systems (including the brain) can implement the necessary inductive biases has been unclear. Here we investigated transitive inference (TI), a classic relational task paradigm in which subjects must learn a relation (A > B and B > C) and generalize it to new combinations of items (A > C). Through mathematical analysis, we found that a broad range of biologically relevant learning models (e.g. gradient flow or ridge regression) perform TI successfully and recapitulate signature behavioral patterns long observed in living subjects. First, we found that models with item-wise additive representations automatically encode transitive relations. Second, for more general representations, a single scalar "conjunctivity factor" determines model behavior on TI and, further, the principle of norm minimization (a standard statistical inductive bias) enables models with fixed, partly conjunctive representations to generalize transitively. Finally, neural networks in the "rich regime," which enables representation learning and has been found to improve generalization, unexpectedly show poor generalization and anomalous behavior. We find that such networks implement a form of norm minimization (over hidden weights) that yields a local encoding mechanism lacking transitivity. Our findings show how minimal statistical learning principles give rise to a classical relational inductive bias (transitivity), explain empirically observed behaviors, and establish a formal approach to understanding the neural basis of relational abstraction.

What does the mean mean? A simple test for neuroscience

Trial-averaged metrics, e.g. tuning curves or population response vectors, are a ubiquitous way of characterizing neuronal activity. But how relevant are such trial-averaged responses to neuronal computation itself? Here we present a simple test to estimate whether average responses reflect aspects of neuronal activity that contribute to neuronal processing. The test probes two assumptions implicitly made whenever average metrics are treated as meaningful representations of neuronal activity: Reliability: Neuronal responses repeat consistently enough across trials that they convey a recognizable reflection of the average response to downstream regions.Behavioural relevance: If a single-trial response is more similar to the average template, it is more likely to evoke correct behavioural responses. We apply this test to two data sets: (1) Two-photon recordings in primary somatosensory cortices (S1 and S2) of mice trained to detect optogenetic stimulation in S1; and (2) Electrophysiological recordings from 71 brain areas in mice performing a contrast discrimination task. Under the highly controlled settings of Data set 1, both assumptions were largely fulfilled. In contrast, the less restrictive paradigm of Data set 2 met neither assumption. Simulations predict that the larger diversity of neuronal response preferences, rather than higher cross-trial reliability, drives the better performance of Data set 1. We conclude that when behaviour is less tightly restricted, average responses do not seem particularly relevant to neuronal computation, potentially because information is encoded more dynamically. Most importantly, we encourage researchers to apply this simple test of computational relevance whenever using trial-averaged neuronal metrics, in order to gauge how representative cross-trial averages are in a given context.

Rationality, preferences, and emotions with biological constraints: it all starts from our senses

Is the role of our sensory systems to represent the physical world as accurately as possible? If so, are our preferences and emotions, often deemed irrational, decoupled from these 'ground-truth' sensory experiences? We show why the answer to both questions is 'no'. Brain function is metabolically costly, and the brain loses some fraction of the information that it encodes and transmits. Therefore, if brains maximize objective functions that increase the fitness of their species, they should adapt to the objective-maximizing rules of the environment at the earliest stages of sensory processing. Consequently, observed 'irrationalities', preferences, and emotions stem from the necessity for our early sensory systems to adapt and process information while considering the metabolic costs and internal states of the organism.

Disinhibition by VIP interneurons is orthogonal to cross-modal attentional modulation in primary visual cortex

Attentional modulation of sensory processing is a key feature of cognition; however, its neural circuit basis is poorly understood. A candidate mechanism is the disinhibition of pyramidal cells through vasoactive intestinal peptide (VIP) and somatostatin (SOM)-positive interneurons. However, the interaction of attentional modulation and VIP-SOM disinhibition has never been directly tested. We used all-optical methods to bi-directionally manipulate VIP interneuron activity as mice performed a cross-modal attention-switching task. We measured the activities of VIP, SOM, and parvalbumin (PV)-positive interneurons and pyramidal neurons identified in the same tissue and found that although activity in all cell classes was modulated by both attention and VIP manipulation, their effects were orthogonal. Attention and VIP-SOM disinhibition relied on distinct patterns of changes in activity and reorganization of interactions between inhibitory and excitatory cells. Circuit modeling revealed a precise network architecture consistent with multiplexing strong yet non-interacting modulations in the same neural population.

A consistent map in the medial entorhinal cortex supports spatial memory

The medial entorhinal cortex (MEC) is hypothesized to function as a cognitive map for memory-guided navigation. How this map develops during learning and influences memory remains unclear. By imaging MEC calcium dynamics while mice successfully learned a novel virtual environment over ten days, we discovered that the dynamics gradually became more spatially consistent and then stabilized. Additionally, grid cells in the MEC not only exhibited improved spatial tuning consistency, but also maintained stable phase relationships, suggesting a network mechanism involving synaptic plasticity and rigid recurrent connectivity to shape grid cell activity during learning. Increased c-Fos expression in the MEC in novel environments further supports the induction of synaptic plasticity. Unsuccessful learning lacked these activity features, indicating that a consistent map is specific for effective spatial memory. Finally, optogenetically disrupting spatial consistency of the map impaired memory-guided navigation in a well-learned environment. Thus, we demonstrate that the establishment of a spatially consistent MEC map across learning both correlates with, and is necessary for, successful spatial memory.

Simultaneous interneuron labeling reveals population-level interactions among parvalbumin, somatostatin, and pyramidal neurons in cortex

Cortical interneurons shape network activity in cell type-specific ways, and are also influenced by interactions with other cell types. These specific cell-type interactions are understudied, as transgenic labeling methods typically restrict labeling to one neuron type at a time. Although recent methods have enabled post-hoc identification of cell types, these are not available to many labs. Here, we present a method to distinguish between two red fluorophores in vivo, which allowed imaging of activity in somatostatin (SOM), parvalbumin (PV), and putative pyramidal neurons (PYR) in mouse association cortex. We compared population events of elevated activity and observed that the PYR network state corresponded to the ratio between mean SOM and PV neuron activity, demonstrating the importance of simultaneous labeling to explain dynamics. These results extend previous findings in sensory cortex, as activity became sparser and less correlated when the ratio between SOM and PV activity was high.

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.

Inferring neural activity before plasticity as a foundation for learning beyond backpropagation

For both humans and machines, the essence of learning is to pinpoint which components in its information processing pipeline are responsible for an error in its output, a challenge that is known as 'credit assignment'. It has long been assumed that credit assignment is best solved by backpropagation, which is also the foundation of modern machine learning. Here, we set out a fundamentally different principle on credit assignment called 'prospective configuration'. In prospective configuration, the network first infers the pattern of neural activity that should result from learning, and then the synaptic weights are modified to consolidate the change in neural activity. We demonstrate that this distinct mechanism, in contrast to backpropagation, (1) underlies learning in a well-established family of models of cortical circuits, (2) enables learning that is more efficient and effective in many contexts faced by biological organisms and (3) reproduces surprising patterns of neural activity and behavior observed in diverse human and rat learning experiments.

Mice Recognise Mice in Neighbouring Rearing Cages and Change Their Social Behaviour

Mice are social animals that change their behaviour primarily in response to visual, olfactory, and auditory information from conspecifics. Rearing conditions such as cage size and colour are important factors influencing mouse behaviour. In recent years, transparent plastic cages have become standard breeding cages. The advantage of using a transparent cage is that the experimenter can observe the mouse from outside the cage without touching the cage. However, mice may recognise the environment outside the cage and change their behaviour. We speculated that mice housed in transparent cages might recognise mice in neighbouring cages. We used only male mice in this experiment. C57BL/6 mice were kept in transparent rearing cages with open lids, and the cage positions were maintained for 3 weeks. Subsequently, we examined how mice behaved toward cagemate mice, mice from neighbouring cages, and mice from distant cages. We compared the level of interest in mice using a social preference test. Similar to previous reports, subject mice showed a high degree of interest in unfamiliar mice from distant cages. By contrast, subject mice reacted to mice from neighbouring cages as familiar mice, similar to cagemate mice. This suggests that mice housed in transparent cages with open lids perceive the external environment and identify mice in neighbouring cages. Researchers should pay attention to the environment outside the mouse cage, especially for the social preference test.

Rapid fluctuations in functional connectivity of cortical networks encode spontaneous behavior

Experimental work across species has demonstrated that spontaneously generated behaviors are robustly coupled to variations in neural activity within the cerebral cortex. Functional magnetic resonance imaging data suggest that temporal correlations in cortical networks vary across distinct behavioral states, providing for the dynamic reorganization of patterned activity. However, these data generally lack the temporal resolution to establish links between cortical signals and the continuously varying fluctuations in spontaneous behavior observed in awake animals. Here, we used wide-field mesoscopic calcium imaging to monitor cortical dynamics in awake mice and developed an approach to quantify rapidly time-varying functional connectivity. We show that spontaneous behaviors are represented by fast changes in both the magnitude and correlational structure of cortical network activity. Combining mesoscopic imaging with simultaneous cellular-resolution two-photon microscopy demonstrated that correlations among neighboring neurons and between local and large-scale networks also encode behavior. Finally, the dynamic functional connectivity of mesoscale signals revealed subnetworks not predicted by traditional anatomical atlas-based parcellation of the cortex. These results provide new insights into how behavioral information is represented across the neocortex and demonstrate an analytical framework for investigating time-varying functional connectivity in neural networks.

Visual experience induces 4-8 Hz synchrony between V1 and higher-order visual areas

Visual perceptual experience induces persistent 4-8 Hz oscillations in the mouse primary visual cortex (V1), encoding visual familiarity. Recent studies suggest that higher-order visual areas (HVAs) are functionally specialized and segregated into information streams processing distinct visual features. However, whether visual memories are processed and stored within the distinct streams is not understood. We report here that V1 and lateromedial (LM), but not V1 and anterolateral, become more phase synchronized in 4-8 Hz after the entrainment of visual stimulus that maximally induces responses in LM. Directed information analysis reveals changes in the top-down functional connectivity between V1 and HVAs. Optogenetic inactivation of LM reduces post-stimulus oscillation peaks in V1 and impairs visual discrimination behavior. Our results demonstrate that 4-8 Hz familiarity-evoked oscillations are specific for the distinct visual features and are present in the corresponding HVAs, where they may be used for the inter-areal communication with V1 during memory-related behaviors.

A direction-selective cortico-brainstem pathway adaptively modulates innate behaviors

Sensory cortices modulate innate behaviors through corticofugal projections targeting phylogenetically-old brainstem nuclei. However, the principles behind the functional connectivity of these projections remain poorly understood. Here, we show that in mice visual cortical neurons projecting to the optic-tract and dorsal-terminal nuclei (NOT-DTN) possess distinct response properties and anatomical connectivity, supporting the adaption of an essential innate eye movement, the optokinetic reflex (OKR). We find that these corticofugal neurons are enriched in specific visual areas, and they prefer temporo-nasal visual motion, matching the direction bias of downstream NOT-DTN neurons. Remarkably, continuous OKR stimulation selectively enhances the activity of these temporo-nasally biased cortical neurons, which can efficiently promote OKR plasticity. Lastly, we demonstrate that silencing downstream NOT-DTN neurons, which project specifically to the inferior olive-a key structure in oculomotor plasticity, impairs the cortical modulation of OKR and OKR plasticity. Our results unveil a direction-selective cortico-brainstem pathway that adaptively modulates innate behaviors.

Reward Modulates Visual Responses in the Superficial Superior Colliculus of Mice

The processing of sensory input is constantly adapting to behavioral demands and internal states. The drive to obtain reward, e.g., searching for water when thirsty, is a strong behavioral demand and associating the reward with its source, a certain environment or action, is paramount for survival. Here, we show that water reward increases subsequent visual activity in the superficial layers of the superior colliculus (SC), which receive direct input from the retina and belong to the earliest stages of visual processing. We trained mice of either sex to perform a visual decision task and recorded the activity of neurons in the SC using two-photon calcium imaging and high-density electrophysiological recordings. Responses to visual stimuli in around 20% of visually responsive neurons in the superficial SC were affected by reward delivered in the previous trial. Reward mostly increased visual responses independent from modulations due to pupil size changes. The modulation of visual responses by reward could not be explained by movements like licking. It was specific to responses to the following visual stimulus, independent of slow fluctuations in neural activity and independent of how often the stimulus was previously rewarded. Electrophysiological recordings confirmed these results and revealed that reward affected the early phase of the visual response around 80 ms after stimulus onset. Modulation of visual responses by reward, but not pupil size, significantly improved the performance of a population decoder to detect visual stimuli, indicating the relevance of reward modulation for the visual performance of the animal.SIGNIFICANCE STATEMENT To learn which actions lead to food, water, or safety, it is necessary to integrate the receiving of reward with sensory stimuli related to the reward. Cortical stages of sensory processing have been shown to represent stimulus-reward associations. Here, we show, however, that reward influences neurons at a much earlier stage of sensory processing, the superior colliculus (SC), receiving direct input from the retina. Visual responses were increased shortly after the animal received the water reward, which led to an improved stimulus signal in the population of these visual neurons. Reward modulation of early visual responses may thus improve perception of visual environments predictive of reward.

Signatures of task learning in neural representations

While neural plasticity has long been studied as the basis of learning, the growth of large-scale neural recording techniques provides a unique opportunity to study how learning-induced activity changes are coordinated across neurons within the same circuit. These distributed changes can be understood through an evolution of the geometry of neural manifolds and latent dynamics underlying new computations. In parallel, studies of multi-task and continual learning in artificial neural networks hint at a tradeoff between non-interference and compositionality as guiding principles to understand how neural circuits flexibly support multiple behaviors. In this review, we highlight recent findings from both biological and artificial circuits that together form a new framework for understanding task learning at the population level.

Neurons in the primary visual cortex of freely moving rats encode both sensory and non-sensory task variables

Neurons in primary visual cortex (area V1) are strongly driven by both sensory stimuli and non-sensory events. However, although the representation of sensory stimuli has been well characterized, much less is known about the representation of non-sensory events. Here, we characterize the specificity and organization of non-sensory representations in rat V1 during a freely moving visual decision task. We find that single neurons encode diverse combinations of task features simultaneously and across task epochs. Despite heterogeneity at the level of single neuron response patterns, both visual and nonvisual task variables could be reliably decoded from small neural populations (5 to 40 units) throughout a trial. Interestingly, in animals trained to make an auditory decision following passive observation of a visual stimulus, some but not all task features could also be decoded from V1 activity. Our results support the view that even in V1-the earliest stage of the cortical hierarchy-bottom-up sensory information may be combined with top-down non-sensory information in a task-dependent manner.

Repeated passive visual experience modulates spontaneous and non-familiar stimuli-evoked neural activity

Familiarity creates subjective memory of repeated innocuous experiences, reduces neural and behavioral responsiveness to those experiences, and enhances novelty detection. The neural correlates of the internal model of familiarity and the cellular mechanisms of enhanced novelty detection following multi-day repeated passive experience remain elusive. Using the mouse visual cortex as a model system, we test how the repeated passive experience of a 45° orientation-grating stimulus for multiple days alters spontaneous and non-familiar stimuli evoked neural activity in neurons tuned to familiar or non-familiar stimuli. We found that familiarity elicits stimulus competition such that stimulus selectivity reduces in neurons tuned to the familiar 45° stimulus; it increases in those tuned to the 90° stimulus but does not affect neurons tuned to the orthogonal 135° stimulus. Furthermore, neurons tuned to orientations 45° apart from the familiar stimulus dominate local functional connectivity. Interestingly, responsiveness to natural images, which consists of familiar and non-familiar orientations, increases subtly in neurons that exhibit stimulus competition. We also show the similarity between familiar grating stimulus-evoked and spontaneous activity increases, indicative of an internal model of altered experience.

Dynamics and maintenance of categorical responses in primary auditory cortex during task engagement

Grouping sets of sounds into relevant categories is an important cognitive ability that enables the association of stimuli with appropriate goal-directed behavioral responses. In perceptual tasks, the primary auditory cortex (A1) assumes a prominent role by concurrently encoding both sound sensory features and task-related variables. Here, we sought to explore the role of A1 in the initiation of sound categorization, shedding light on its involvement in this cognitive process. We trained ferrets to discriminate click trains of different rates in a Go/No-Go delayed categorization task and recorded neural activity during both active behavior and passive exposure to the same sounds. Purely categorical response components were extracted and analyzed separately from sensory responses to reveal their contributions to the overall population response throughout the trials. We found that categorical activity emerged during sound presentation in the population average and was present in both active behavioral and passive states. However, upon task engagement, categorical responses to the No-Go category became suppressed in the population code, leading to an asymmetrical representation of the Go stimuli relative to the No-Go sounds and pre-stimulus baseline. The population code underwent an abrupt change at stimulus offset, with sustained responses after the Go sounds during the delay period. Notably, the categorical responses observed during the stimulus period exhibited a significant correlation with those extracted from the delay epoch, suggesting an early involvement of A1 in stimulus categorization.

History information emerges in the cortex during learning

We learn from our experience but the underlying neuronal mechanisms incorporating past information to facilitate learning is relatively unknown. Specifically, which cortical areas encode history-related information and how is this information modulated across learning? To study the relationship between history and learning, we continuously imaged cortex-wide calcium dynamics as mice learn to use their whiskers to discriminate between two different textures. We mainly focused on comparing the same trial type with different trial history, that is, a different preceding trial. We found trial history information in barrel cortex (BC) during stimulus presentation. Importantly, trial history in BC emerged only as the mouse learned the task. Next, we also found learning-dependent trial history information in rostrolateral (RL) association cortex that emerges before stimulus presentation, preceding activity in BC. Trial history was also encoded in other cortical areas and was not related to differences in body movements. Interestingly, a binary classifier could discriminate trial history at the single trial level just as well as current information both in BC and RL. These findings suggest that past experience emerges in the cortex around the time of learning, starting from higher-order association area RL and propagating down (i.e., top-down projection) to lower-order BC where it can be integrated with incoming sensory information. This integration between the past and present may facilitate learning.

Distinct Patterns of Connectivity between Brain Regions Underlie the Intra-Modal and Cross-Modal Value-Driven Modulations of the Visual Cortex

Past reward associations may be signaled from different sensory modalities; however, it remains unclear how different types of reward-associated stimuli modulate sensory perception. In this human fMRI study (female and male participants), a visual target was simultaneously presented with either an intra- (visual) or a cross-modal (auditory) cue that was previously associated with rewards. We hypothesized that, depending on the sensory modality of the cues, distinct neural mechanisms underlie the value-driven modulation of visual processing. Using a multivariate approach, we confirmed that reward-associated cues enhanced the target representation in early visual areas and identified the brain valuation regions. Then, using an effective connectivity analysis, we tested three possible patterns of connectivity that could underlie the modulation of the visual cortex: a direct pathway from the frontal valuation areas to the visual areas, a mediated pathway through the attention-related areas, and a mediated pathway that additionally involved sensory association areas. We found evidence for the third model demonstrating that the reward-related information in both sensory modalities is communicated across the valuation and attention-related brain regions. Additionally, the superior temporal areas were recruited when reward was cued cross-modally. The strongest dissociation between the intra- and cross-modal reward-driven effects was observed at the level of the feedforward and feedback connections of the visual cortex estimated from the winning model. These results suggest that, in the presence of previously rewarded stimuli from different sensory modalities, a combination of domain-general and domain-specific mechanisms are recruited across the brain to adjust the visual perception.SIGNIFICANCE STATEMENT Reward has a profound effect on perception, but it is not known whether shared or disparate mechanisms underlie the reward-driven effects across sensory modalities. In this human fMRI study, we examined the reward-driven modulation of the visual cortex by visual (intra-modal) and auditory (cross-modal) reward-associated cues. Using a model-based approach to identify the most plausible pattern of inter-regional effective connectivity, we found that higher-order areas involved in the valuation and attentional processing were recruited by both types of rewards. However, the pattern of connectivity between these areas and the early visual cortex was distinct between the intra- and cross-modal rewards. This evidence suggests that, to effectively adapt to the environment, reward signals may recruit both domain-general and domain-specific mechanisms.

Electrophysiological Signatures of Visual Recognition Memory across All Layers of Mouse V1

In mouse primary visual cortex (V1), familiar stimuli evoke significantly altered responses when compared with novel stimuli. This stimulus-selective response plasticity (SRP) was described originally as an increase in the magnitude of visual evoked potentials (VEPs) elicited in layer 4 (L4) by familiar phase-reversing grating stimuli. SRP is dependent on NMDA receptors (NMDARs) and has been hypothesized to reflect potentiation of thalamocortical (TC) synapses in L4. However, recent evidence indicates that the synaptic modifications that manifest as SRP do not occur on L4 principal cells. To shed light on where and how SRP is induced and expressed in male and female mice, the present study had three related aims: (1) to confirm that NMDAR are required specifically in glutamatergic principal neurons of V1, (2) to investigate the consequences of deleting NMDAR specifically in L6, and (3) to use translaminar electrophysiological recordings to characterize SRP expression in different layers of V1. We find that knock-out (KO) of NMDAR in L6 principal neurons disrupts SRP. Current-source density (CSD) analysis of the VEP depth profile shows augmentation of short latency current sinks in layers 3, 4, and 6 in response to phase reversals of familiar stimuli. Multiunit recordings demonstrate that increased peak firing occurs in response to phase reversals of familiar stimuli across all layers, but that activity between phase reversals is suppressed. Together, these data reveal important aspects of the underlying phenomenology of SRP and generate new hypotheses for the expression of experience-dependent plasticity in V1.SIGNIFICANCE STATEMENT Repeated exposure to stimuli that portend neither reward nor punishment leads to behavioral habituation, enabling organisms to dedicate attention to novel or otherwise significant features of the environment. The neural basis of this process, which is so often dysregulated in neurologic and psychiatric disorders, remains poorly understood. Learning and memory of stimulus familiarity can be studied in mouse visual cortex by measuring electrophysiological responses to simple phase-reversing grating stimuli. The current study advances knowledge of this process by documenting changes in visual evoked potentials (VEPs), neuronal spiking activity, and oscillations in the local field potentials (LFPs) across all layers of mouse visual cortex. In addition, we identify a key contribution of a specific population of neurons in layer 6 (L6) of visual cortex.

Continuous multiplexed population representations of task context in the mouse primary visual cortex

Effective task execution requires the representation of multiple task-related variables that determine how stimuli lead to correct responses. Even the primary visual cortex (V1) represents other task-related variables such as expectations, choice, and context. However, it is unclear how V1 can flexibly accommodate these variables without interfering with visual representations. We trained mice on a context-switching cross-modal decision task, where performance depends on inferring task context. We found that the context signal that emerged in V1 was behaviorally relevant as it strongly covaried with performance, independent from movement. Importantly, this signal was integrated into V1 representation by multiplexing visual and context signals into orthogonal subspaces. In addition, auditory and choice signals were also multiplexed as these signals were orthogonal to the context representation. Thus, multiplexing allows V1 to integrate visual inputs with other sensory modalities and cognitive variables to avoid interference with the visual representation while ensuring the maintenance of task-relevant variables.

A consistent map in the medial entorhinal cortex supports spatial memory

The medial entorhinal cortex (MEC) is hypothesized to function as a cognitive map for memory-guided navigation. How this map develops during learning and influences memory remains unclear. By imaging MEC calcium dynamics while mice successfully learned a novel virtual environment over ten days, we discovered that the dynamics gradually became more spatially consistent and then stabilized. Additionally, grid cells in the MEC not only exhibited improved spatial tuning consistency, but also maintained stable phase relationships, suggesting a network mechanism involving synaptic plasticity and rigid recurrent connectivity to shape grid cell activity during learning. Increased c-Fos expression in the MEC in novel environments further supports the induction of synaptic plasticity. Unsuccessful learning lacked these activity features, indicating that a consistent map is specific for effective spatial memory. Finally, optogenetically disrupting spatial consistency of the map impaired memory-guided navigation in a well-learned environment. Thus, we demonstrate that the establishment of a spatially consistent MEC map across learning both correlates with, and is necessary for, successful spatial memory.

Learning induced neuronal identity switch in the superficial layers of the primary somatosensory cortex

During learning, multi-dimensional inputs are integrated within the sensory cortices. However, the strategies by which the sensory cortex employs to achieve learning remains poorly understood. We studied the sensory cortical neuronal coding of trace eyeblink conditioning (TEC) in head-fixed, freely running mice, where whisker deflection was used as a conditioned stimulus (CS) and an air puff to the cornea delivered after an interval was used as unconditioned stimulus (US). After training, mice learned the task with a set of stereotypical behavioral changes, most prominent ones include prolonged closure of eyelids, and increased reverse running between CS and US onset. The local blockade of the primary somatosensory cortex (S1) activities with muscimol abolished the behavior learning suggesting that S1 is required for the TEC. In naive animals, based on the response properties to the CS and US, identities of the small proportion (~20%) of responsive primary neurons (PNs) were divided into two subtypes: CR (i.e. CS-responsive) and UR neurons (i.e. US-responsive). After animals learned the task, identity of CR and UR neurons changed: while the CR neurons are less responsive to CS, UR neurons gain responsiveness to CS, a new phenomenon we defined as 'learning induced neuronal identity switch (LINIS)'. To explore the potential mechanisms underlying LINIS, we found that systemic and local (i.e. in S1) administration of the nicotinic receptor antagonist during TEC training blocked the LINIS, and concomitantly disrupted the behavior learning. Additionally, we monitored responses of two types of cortical interneurons (INs) and observed that the responses of the somatostatin-expressing (SST), but not parvalbumin-expressing (PV) INs are negatively correlated with the learning performance, suggesting that SST-INs contribute to the LINIS. Thus, we conclude that L2/3 PNs in S1 encode perceptual learning by LINIS like mechanisms, and cholinergic pathways and cortical SST interneurons are involved in the formation of LINIS.

Repeated passive visual experience modulates spontaneous and non-familiar stimulievoked neural activity

Familiarity creates subjective memory of repeated innocuous experiences, reduces neural and behavioral responsiveness to those experiences, and enhances novelty detection. The neural correlates of the internal model of familiarity and the cellular mechanisms of enhanced novelty detection following multi-day repeated passive experience remain elusive. Using the mouse visual cortex as a model system, we test how the repeated passive experience of a 45° orientation-grating stimulus for multiple days alters spontaneous and non-familiar stimuli evoked neural activity in neurons tuned to familiar or non-familiar stimuli. We found that familiarity elicits stimulus competition such that stimulus selectivity reduces in neurons tuned to the familiar 45° stimulus; it increases in those tuned to the 90° stimulus but does not affect neurons tuned to the orthogonal 135° stimulus. Furthermore, neurons tuned to orientations 45° apart from the familiar stimulus dominate local functional connectivity. Interestingly, responsiveness to natural images, which consists of familiar and non-familiar orientations, increases subtly in neurons that exhibit stimulus competition. We also show the similarity between familiar grating stimulus-evoked and spontaneous activity increases, indicative of an internal model of altered experience.

Lexical and sublexical cortical tuning for print revealed by Steady-State Visual Evoked Potentials (SSVEPs) in early readers

There are multiple levels of processing relevant to reading that vary in their visual, sublexical, and lexical orthographic processing demands. Segregating distinct cortical sources for each of these levels has been challenging in EEG studies of early readers. To address this challenge, we applied recent advances in analyzing high-density EEG using Steady-State Visual Evoked Potentials (SSVEPs) via data-driven Reliable Components Analysis (RCA) in a group of early readers spanning from kindergarten to second grade. Three controlled stimulus contrasts-familiar words versus unfamiliar pseudofonts, familiar words versus pseudowords, and pseudowords versus nonwords-were used to isolate coarse print tuning, lexical processing, and sublexical orthography-related processing, respectively. First, three overlapping yet distinct neural sources-left vOT, dorsal parietal, and primary visual cortex were revealed underlying coarse print tuning. Second, we segregated distinct cortical sources for the other two levels of processing: lexical fine tuning over occipito-temporal/parietal regions; sublexical orthographic fine tuning over left occipital regions. Finally, exploratory group analyses based on children's reading fluency suggested that coarse print tuning emerges early even in children with limited reading knowledge, while sublexical and higher-level lexical processing emerge only in children with sufficient reading knowledge. RESEARCH HIGHLIGHTS: Cognitive processes underlying coarse print tuning, sublexical, and lexical fine tuning were examined in beginning readers. Three overlapping yet distinct neural sources-left ventral occipito-temporal (vOT), left temporo-parietal, and primary visual cortex-were revealed underlying coarse print tuning. Responses to sublexical orthographic fine tuning were found over left occipital regions, while responses to higher-level linguistic fine tuning were found over occipito-temporal/parietal regions. Exploratory group analyses suggested that coarse print tuning emerges in children with limited reading knowledge, while sublexical and higher-level linguistic fine tuning effects emerge in children with sufficient reading knowledge.

Emerging V1 neuronal ensembles with enhanced connectivity after associative learning

Introduction

The visual stimulus-specific responses in the primary visual cortex (V1) undergo plastic changes after associative learning. During the learning process, neuronal ensembles, defined as groups of coactive neurons, are well known to be related to learning and memory. However, it remains unclear what effect learning has on ensembles, and which neuronal subgroups within those ensembles play a key role in associative learning.

Methods

We used two-photon calcium imaging in mice to record the activity of V1 neurons before and after fear conditioning associated with a visual cue (blue light). We first defined neuronal ensembles by thresholding their functional connectivity in response to blue (conditioned) or green (control) light. We defined neurons that existed both before and after conditioning as stable neurons. Neurons which were recruited after conditioning were defined as new neurons. The graph theory-based analysis was performed to quantify the changes in connectivity within ensembles after conditioning.

Results

A significant enhancement in the connectivity strength (the average correlation with other neurons) was observed in the blue ensembles after conditioning. We found that stable neurons within the blue ensembles showed a significantly smaller clustering coefficient (the value represented the degree of interconnectedness among a node's neighbors) after conditioning than they were before conditioning. Additionally, new neurons within the blue ensembles had a larger clustering coefficient, similar relative degree (the value represented the number of functional connections between neurons) and connectivity strength compared to stable neurons in the same ensembles.

Discussion

Overall, our results demonstrated that the plastic changes caused by conditioning occurred in subgroups of neurons in the ensembles. Moreover, new neurons from conditioned ensembles may play a crucial role in memory formation, as they exhibited not only similar connection competence in relative degree and connectivity strength as stable neurons, but also showed a significantly larger clustering coefficient compared to the stable neurons within the same ensembles after conditioning.

Learning in a sensory cortical microstimulation task is associated with elevated representational stability

Sensory cortical representations can be highly dynamic, raising the question of how representational stability impacts learning. We train mice to discriminate the number of photostimulation pulses delivered to opsin-expressing pyramidal neurons in layer 2/3 of primary vibrissal somatosensory cortex. We simultaneously track evoked neural activity across learning using volumetric two-photon calcium imaging. In well-trained animals, trial-to-trial fluctuations in the amount of photostimulus-evoked activity predicted animal choice. Population activity levels declined rapidly across training, with the most active neurons showing the largest declines in responsiveness. Mice learned at varied rates, with some failing to learn the task in the time provided. The photoresponsive population showed greater instability both within and across behavioral sessions among animals that failed to learn. Animals that failed to learn also exhibited a faster deterioration in stimulus decoding. Thus, greater stability in the stimulus response is associated with learning in a sensory cortical microstimulation task.

Human orbitofrontal cortex signals decision outcomes to sensory cortex during behavioral adaptations

The ability to respond flexibly to an ever-changing environment relies on the orbitofrontal cortex (OFC). However, how the OFC associates sensory information with predicted outcomes to enable flexible sensory learning in humans remains elusive. Here, we combine a probabilistic tactile reversal learning task with functional magnetic resonance imaging (fMRI) to investigate how lateral OFC (lOFC) interacts with the primary somatosensory cortex (S1) to guide flexible tactile learning in humans. fMRI results reveal that lOFC and S1 exhibit distinct task-dependent engagement: while the lOFC responds transiently to unexpected outcomes immediately following reversals, S1 is persistently engaged during re-learning. Unlike the contralateral stimulus-selective S1, activity in ipsilateral S1 mirrors the outcomes of behavior during re-learning, closely related to top-down signals from lOFC. These findings suggest that lOFC contributes to teaching signals to dynamically update representations in sensory areas, which implement computations critical for adaptive behavior.

Visual cortex: How mice learn to detect entirely novel inputs

A new study has revealed that neural amplification in mouse primary visual cortex substantially increases between training sessions as mice learn to detect novel optogenetic input directly into visual cortex, suggesting consolidation and recurrent network plasticity contribute to learning the behavior.

Differential encoding of temporal context and expectation under representational drift across hierarchically connected areas

The classic view that neural populations in sensory cortices preferentially encode responses to incoming stimuli has been strongly challenged by recent experimental studies. Despite the fact that a large fraction of variance of visual responses in rodents can be attributed to behavioral state and movements, trial-history, and salience, the effects of contextual modulations and expectations on sensory-evoked responses in visual and association areas remain elusive. Here, we present a comprehensive experimental and theoretical study showing that hierarchically connected visual and association areas differentially encode the temporal context and expectation of naturalistic visual stimuli, consistent with the theory of hierarchical predictive coding. We measured neural responses to expected and unexpected sequences of natural scenes in the primary visual cortex (V1), the posterior medial higher order visual area (PM), and retrosplenial cortex (RSP) using 2-photon imaging in behaving mice collected through the Allen Institute Mindscope's OpenScope program. We found that information about image identity in neural population activity depended on the temporal context of transitions preceding each scene, and decreased along the hierarchy. Furthermore, our analyses revealed that the conjunctive encoding of temporal context and image identity was modulated by expectations of sequential events. In V1 and PM, we found enhanced and specific responses to unexpected oddball images, signaling stimulus-specific expectation violation. In contrast, in RSP the population response to oddball presentation recapitulated the missing expected image rather than the oddball image. These differential responses along the hierarchy are consistent with classic theories of hierarchical predictive coding whereby higher areas encode predictions and lower areas encode deviations from expectation. We further found evidence for drift in visual responses on the timescale of minutes. Although activity drift was present in all areas, population responses in V1 and PM, but not in RSP, maintained stable encoding of visual information and representational geometry. Instead we found that RSP drift was independent of stimulus information, suggesting a role in generating an internal model of the environment in the temporal domain. Overall, our results establish temporal context and expectation as substantial encoding dimensions in the visual cortex subject to fast representational drift and suggest that hierarchically connected areas instantiate a predictive coding mechanism.

Amplified cortical neural responses as animals learn to use novel activity patterns

Cerebral cortex supports representations of the world in patterns of neural activity, used by the brain to make decisions and guide behavior. Past work has found diverse, or limited, changes in the primary sensory cortex in response to learning, suggesting that the key computations might occur in downstream regions. Alternatively, sensory cortical changes may be central to learning. We studied cortical learning by using controlled inputs we insert: we trained mice to recognize entirely novel, non-sensory patterns of cortical activity in the primary visual cortex (V1) created by optogenetic stimulation. As animals learned to use these novel patterns, we found that their detection abilities improved by an order of magnitude or more. The behavioral change was accompanied by large increases in V1 neural responses to fixed optogenetic input. Neural response amplification to novel optogenetic inputs had little effect on existing visual sensory responses. A recurrent cortical model shows that this amplification can be achieved by a small mean shift in recurrent network synaptic strength. Amplification would seem to be desirable to improve decision-making in a detection task; therefore, these results suggest that adult recurrent cortical plasticity plays a significant role in improving behavioral performance during learning.

Sensory perception relies on fitness-maximizing codes

Sensory information encoded by humans and other organisms is generally presumed to be as accurate as their biological limitations allow. However, perhaps counterintuitively, accurate sensory representations may not necessarily maximize the organism's chances of survival. To test this hypothesis, we developed a unified normative framework for fitness-maximizing encoding by combining theoretical insights from neuroscience, computer science, and economics. Behavioural experiments in humans revealed that sensory encoding strategies are flexibly adapted to promote fitness maximization, a result confirmed by deep neural networks with information capacity constraints trained to solve the same task as humans. Moreover, human functional MRI data revealed that novel behavioural goals that rely on object perception induce efficient stimulus representations in early sensory structures. These results suggest that fitness-maximizing rules imposed by the environment are applied at early stages of sensory processing in humans and machines.

A distributed and efficient population code of mixed selectivity neurons for flexible navigation decisions

Decision-making requires flexibility to rapidly switch one's actions in response to sensory stimuli depending on information stored in memory. We identified cortical areas and neural activity patterns underlying this flexibility during virtual navigation, where mice switched navigation toward or away from a visual cue depending on its match to a remembered cue. Optogenetics screening identified V1, posterior parietal cortex (PPC), and retrosplenial cortex (RSC) as necessary for accurate decisions. Calcium imaging revealed neurons that can mediate rapid navigation switches by encoding a mixture of a current and remembered visual cue. These mixed selectivity neurons emerged through task learning and predicted the mouse's choices by forming efficient population codes before correct, but not incorrect, choices. They were distributed across posterior cortex, even V1, and were densest in RSC and sparsest in PPC. We propose flexibility in navigation decisions arises from neurons that mix visual and memory information within a visual-parietal-retrosplenial network.

Synthetic Data Resource and Benchmarks for Time Cell Analysis and Detection Algorithms

Hippocampal CA1 cells take part in reliable, time-locked activity sequences in tasks that involve an association between temporally separated stimuli, in a manner that tiles the interval between the stimuli. Such cells have been termed time cells. Here, we adopt a first-principles approach to comparing diverse analysis and detection algorithms for identifying time cells. We generated synthetic activity datasets using calcium signals recorded in vivo from the mouse hippocampus using two-photon (2-P) imaging, as template response waveforms. We assigned known, ground truth values to perturbations applied to perfect activity signals, including noise, calcium event width, timing imprecision, hit trial ratio and background (untuned) activity. We tested a range of published and new algorithms and their variants on this dataset. We find that most algorithms correctly classify over 80% of cells, but have different balances between true and false positives, and different sensitivity to the five categories of perturbation. Reassuringly, most methods are reasonably robust to perturbations, including background activity, and show good concordance in classification of time cells. The same algorithms were also used to analyze and identify time cells in experimental physiology datasets recorded in vivo and most show good concordance.

Cocaine regulates sensory filtering in cortical pyramidal neurons

Exposure to cocaine leads to robust changes in the structure and function of neurons within the mesocorticolimbic pathway. However, little is known about how cocaine influences the processing of information within the sensory cortex. We address this by using patch-clamp and juxtacellular voltage recordings and two-photon Ca2+ imaging in vivo to investigate the influence of acute cocaine exposure on layer 2/3 (L2/3) pyramidal neurons within the primary somatosensory cortex (S1). Here, cocaine dampens membrane potential state transitions and decreases spontaneous somatic action potentials and Ca2+ transients. In contrast to the uniform decrease in background spontaneous activity, cocaine has a heterogeneous influence on sensory encoding, increasing tactile-evoked responses in dendrites that do not typically encode sensory information and decreasing responses in those dendrites that are more reliable sensory encoders. Combined, these findings suggest that cocaine acts as a filter that suppresses background noise to selectively modulate incoming sensory information.