Feature selectivity can explain mismatch signals in mouse visual cortex

Flexible neural population dynamics govern the speed and stability of sensory encoding in mouse visual cortex

Time courses of neural responses underlie real-time sensory processing and perception. How these temporal dynamics change may be fundamental to how sensory systems adapt to different perceptual demands. By simultaneously recording from hundreds of neurons in mouse primary visual cortex, we examined neural population responses to visual stimuli at sub-second timescales, during different behavioural states. We discovered that during active behavioural states characterised by locomotion, single-neurons shift from transient to sustained response modes, facilitating rapid emergence of visual stimulus tuning. Differences in single-neuron response dynamics were associated with changes in temporal dynamics of neural correlations, including faster stabilisation of stimulus-evoked changes in the structure of correlations during locomotion. Using Factor Analysis, we examined temporal dynamics of latent population responses and discovered that trajectories of population activity make more direct transitions between baseline and stimulus-encoding neural states during locomotion. This could be partly explained by dampening of oscillatory dynamics present during stationary behavioural states. Functionally, changes in temporal response dynamics collectively enabled faster, more stable and more efficient encoding of new visual information during locomotion. These findings reveal a principle of how sensory systems adapt to perceptual demands, where flexible neural population dynamics govern the speed and stability of sensory encoding.

Modeling circuit mechanisms of opposing cortical responses to visual flow perturbations

In an ever-changing visual world, animals' survival depends on their ability to perceive and respond to rapidly changing motion cues. The primary visual cortex (V1) is at the forefront of this sensory processing, orchestrating neural responses to perturbations in visual flow. However, the underlying neural mechanisms that lead to distinct cortical responses to such perturbations remain enigmatic. In this study, our objective was to uncover the neural dynamics that govern V1 neurons' responses to visual flow perturbations using a biologically realistic computational model. By subjecting the model to sudden changes in visual input, we observed opposing cortical responses in excitatory layer 2/3 (L2/3) neurons, namely, depolarizing and hyperpolarizing responses. We found that this segregation was primarily driven by the competition between external visual input and recurrent inhibition, particularly within L2/3 and L4. This division was not observed in excitatory L5/6 neurons, suggesting a more prominent role for inhibitory mechanisms in the visual processing of the upper cortical layers. Our findings share similarities with recent experimental studies focusing on the opposing influence of top-down and bottom-up inputs in the mouse primary visual cortex during visual flow perturbations.

Experience shapes chandelier cell function and structure in the visual cortex

Detailed characterization of interneuron types in primary visual cortex (V1) has greatly contributed to understanding visual perception, yet the role of chandelier cells (ChCs) in visual processing remains poorly characterized. Using viral tracing we found that V1 ChCs predominantly receive monosynaptic input from local layer 5 pyramidal cells and higher-order cortical regions. Two-photon calcium imaging and convolutional neural network modeling revealed that ChCs are visually responsive but weakly selective for stimulus content. In mice running in a virtual tunnel, ChCs respond strongly to events known to elicit arousal, including locomotion and visuomotor mismatch. Repeated exposure of the mice to the virtual tunnel was accompanied by reduced visual responses of ChCs and structural plasticity of ChC boutons and axon initial segment length. Finally, ChCs only weakly inhibited pyramidal cells. These findings suggest that ChCs provide an arousal-related signal to layer 2/3 pyramidal cells that may modulate their activity and/or gate plasticity of their axon initial segments during behaviorally relevant events.

Computational components of visual predictive coding circuitry

If a full visual percept can be said to be a 'hypothesis', so too can a neural 'prediction' - although the latter addresses one particular component of image content (such as 3-dimensional organisation, the interplay between lighting and surface colour, the future trajectory of moving objects, and so on). And, because processing is hierarchical, predictions generated at one level are conveyed in a backward direction to a lower level, seeking to predict, in fact, the neural activity at that prior stage of processing, and learning from errors signalled in the opposite direction. This is the essence of 'predictive coding', at once an algorithm for information processing and a theoretical basis for the nature of operations performed by the cerebral cortex. Neural models for the implementation of predictive coding invoke specific functional classes of neuron for generating, transmitting and receiving predictions, and for producing reciprocal error signals. Also a third general class, 'precision' neurons, tasked with regulating the magnitude of error signals contingent upon the confidence placed upon the prediction, i.e., the reliability and behavioural utility of the sensory data that it predicts. So, what is the ultimate source of a 'prediction'? The answer is multifactorial: knowledge of the current environmental context and the immediate past, allied to memory and lifetime experience of the way of the world, doubtless fine-tuned by evolutionary history too. There are, in consequence, numerous potential avenues for experimenters seeking to manipulate subjects' expectation, and examine the neural signals elicited by surprising, and less surprising visual stimuli. This review focuses upon the predictive physiology of mouse and monkey visual cortex, summarising and commenting on evidence to date, and placing it in the context of the broader field. It is concluded that predictive coding has a firm grounding in basic neuroscience and that, unsurprisingly, there remains much to learn.

How deep is the brain? The shallow brain hypothesis

Deep learning and predictive coding architectures commonly assume that inference in neural networks is hierarchical. However, largely neglected in deep learning and predictive coding architectures is the neurobiological evidence that all hierarchical cortical areas, higher or lower, project to and receive signals directly from subcortical areas. Given these neuroanatomical facts, today's dominance of cortico-centric, hierarchical architectures in deep learning and predictive coding networks is highly questionable; such architectures are likely to be missing essential computational principles the brain uses. In this Perspective, we present the shallow brain hypothesis: hierarchical cortical processing is integrated with a massively parallel process to which subcortical areas substantially contribute. This shallow architecture exploits the computational capacity of cortical microcircuits and thalamo-cortical loops that are not included in typical hierarchical deep learning and predictive coding networks. We argue that the shallow brain architecture provides several critical benefits over deep hierarchical structures and a more complete depiction of how mammalian brains achieve fast and flexible computational capabilities.

Stimulus-Specific Prediction Error Neurons in Mouse Auditory Cortex

Comparing expectation with experience is an important neural computation performed throughout the brain and is a hallmark of predictive processing. Experiments that alter the sensory outcome of an animal's behavior reveal enhanced neural responses to unexpected self-generated stimuli, indicating that populations of neurons in sensory cortex may reflect prediction errors (PEs), mismatches between expectation and experience. However, enhanced neural responses to self-generated stimuli could also arise through nonpredictive mechanisms, such as the movement-based facilitation of a neuron's inherent sound responses. If sensory prediction error neurons exist in sensory cortex, it is unknown whether they manifest as general error responses, or respond with specificity to errors in distinct stimulus dimensions. To answer these questions, we trained mice of either sex to expect the outcome of a simple sound-generating behavior and recorded auditory cortex activity as mice heard either the expected sound or sounds that deviated from expectation in one of multiple distinct dimensions. Our data reveal that the auditory cortex learns to suppress responses to self-generated sounds along multiple acoustic dimensions simultaneously. We identify a distinct population of auditory cortex neurons that are not responsive to passive sounds or to the expected sound but that encode prediction errors. These prediction error neurons are abundant only in animals with a learned motor-sensory expectation, and encode one or two specific violations rather than a generic error signal. Together, these findings reveal that cortical predictions about self-generated sounds have specificity in multiple simultaneous dimensions and that cortical prediction error neurons encode specific violations from expectation.SIGNIFICANCE STATEMENT Audette et. al record neural activity in the auditory cortex while mice perform a sound-generating forelimb movement and measure neural responses to sounds that violate an animal's expectation in different ways. They find that predictions about self-generated sounds are highly specific across multiple stimulus dimensions and that a population of typically nonsound-responsive neurons respond to sounds that violate an animal's expectation in a specific way. These results identify specific prediction error (PE) signals in the mouse auditory cortex and suggest that errors may be calculated early in sensory processing.

Sharing neurophysiology data from the Allen Brain Observatory

Nullius in verba ('trust no one'), chosen as the motto of the Royal Society in 1660, implies that independently verifiable observations-rather than authoritative claims-are a defining feature of empirical science. As the complexity of modern scientific instrumentation has made exact replications prohibitive, sharing data is now essential for ensuring the trustworthiness of one's findings. While embraced in spirit by many, in practice open data sharing remains the exception in contemporary systems neuroscience. Here, we take stock of the Allen Brain Observatory, an effort to share data and metadata associated with surveys of neuronal activity in the visual system of laboratory mice. Data from these surveys have been used to produce new discoveries, to validate computational algorithms, and as a benchmark for comparison with other data, resulting in over 100 publications and preprints to date. We distill some of the lessons learned about open surveys and data reuse, including remaining barriers to data sharing and what might be done to address these.

Stimulus-specific prediction error neurons in mouse auditory cortex

Comparing expectation with experience is an important neural computation performed throughout the brain and is a hallmark of predictive processing. Experiments that alter the sensory outcome of an animal's behavior reveal enhanced neural responses to unexpected self-generated stimuli, indicating that populations of neurons in sensory cortex may reflect prediction errors - mismatches between expectation and experience. However, enhanced neural responses to self-generated stimuli could also arise through non-predictive mechanisms, such as the movement-based facilitation of a neuron's inherent sound responses. If sensory prediction error neurons exist in sensory cortex, it is unknown whether they manifest as general error responses, or respond with specificity to errors in distinct stimulus dimensions. To answer these questions, we trained mice to expect the outcome of a simple sound-generating behavior and recorded auditory cortex activity as mice heard either the expected sound or sounds that deviated from expectation in one of multiple distinct dimensions. Our data reveal that the auditory cortex learns to suppress responses to self-generated sounds along multiple acoustic dimensions simultaneously. We identify a distinct population of auditory cortex neurons that are not responsive to passive sounds or to the expected sound but that explicitly encode prediction errors. These prediction error neurons are abundant only in animals with a learned motor-sensory expectation, and encode one or two specific violations rather than a generic error signal.

Locomotion-induced gain of visual responses cannot explain visuomotor mismatch responses in layer 2/3 of primary visual cortex

The aim of this work is to provide a comment on a recent paper by Muzzu and Saleem (2021), which claims that visuomotor mismatch responses in mouse visual cortex can be explained by a locomotion-induced gain of visual halt responses. Our primary concern is that without directly comparing these responses with mismatch responses, the claim that one response can explain the other appears difficult to uphold, more so because previous work finds that a uniform locomotion-induced gain cannot explain mismatch responses. To support these arguments, we analyze layer 2/3 calcium imaging datasets and show that coupling between visual flow and locomotion greatly enhances mismatch responses in an experience-dependent manner compared with halts in non-coupled visual flow. This is consistent with mismatch responses representing visuomotor prediction errors. Thus, we conclude that while feature selectivity might contribute to mismatch responses in mouse visual cortex, it cannot explain these responses.

Redefining sensorimotor mismatch selectivity in the visual cortex

This Matters Arising Response contains our commentary to the response written by Vasilevskaya et al., 2023, publishing concurrently in Cell Reports, for our recent article "Feature selectivity can explain mismatch signals in mouse visual cortex." We find that results in the response reinforced many of our findings and, further supported by their new results, we argue for the necessity to redefine sensorimotor mismatch selectivity in the mouse visual system.

Walking humans and running mice: perception and neural encoding of optic flow during self-motion

Locomotion produces full-field optic flow that often dominates the visual motion inputs to an observer. The perception of optic flow is in turn important for animals to guide their heading and interact with moving objects. Understanding how locomotion influences optic flow processing and perception is therefore essential to understand how animals successfully interact with their environment. Here, we review research investigating how perception and neural encoding of optic flow are altered during self-motion, focusing on locomotion. Self-motion has been found to influence estimation and sensitivity for optic flow speed and direction. Nonvisual self-motion signals also increase compensation for self-driven optic flow when parsing the visual motion of moving objects. The integration of visual and nonvisual self-motion signals largely follows principles of Bayesian inference and can improve the precision and accuracy of self-motion perception. The calibration of visual and nonvisual self-motion signals is dynamic, reflecting the changing visuomotor contingencies across different environmental contexts. Throughout this review, we consider experimental research using humans, non-human primates and mice. We highlight experimental challenges and opportunities afforded by each of these species and draw parallels between experimental findings. These findings reveal a profound influence of locomotion on optic flow processing and perception across species. This article is part of a discussion meeting issue 'New approaches to 3D vision'.

Precise movement-based predictions in the mouse auditory cortex

Many of the sensations experienced by an organism are caused by their own actions, and accurately anticipating both the sensory features and timing of self-generated stimuli is crucial to a variety of behaviors. In the auditory cortex, neural responses to self-generated sounds exhibit frequency-specific suppression, suggesting that movement-based predictions may be implemented early in sensory processing. However, it remains unknown whether this modulation results from a behaviorally specific and temporally precise prediction, nor is it known whether corresponding expectation signals are present locally in the auditory cortex. To address these questions, we trained mice to expect the precise acoustic outcome of a forelimb movement using a closed-loop sound-generating lever. Dense neuronal recordings in the auditory cortex revealed suppression of responses to self-generated sounds that was specific to the expected acoustic features, to a precise position within the movement, and to the movement that was coupled to sound during training. Prediction-based suppression was concentrated in L2/3 and L5, where deviations from expectation also recruited a population of prediction-error neurons that was otherwise unresponsive. Recording in the absence of sound revealed abundant movement signals in deep layers that were biased toward neurons tuned to the expected sound, as well as expectation signals that were present throughout the cortex and peaked at the time of expected auditory feedback. Together, these findings identify distinct populations of auditory cortical neurons with movement, expectation, and error signals consistent with a learned internal model linking an action to its specific acoustic outcome.

Learning enhances encoding of time and temporal surprise in mouse primary sensory cortex

Primary sensory cortex has long been believed to play a straightforward role in the initial processing of sensory information. Yet, the superficial layers of cortex overall are sparsely active, even during sensory stimulation; additionally, cortical activity is influenced by other modalities, task context, reward, and behavioral state. Our study demonstrates that reinforcement learning dramatically alters representations among longitudinally imaged neurons in superficial layers of mouse primary somatosensory cortex. Learning an object detection task recruits previously unresponsive neurons, enlarging the neuronal population sensitive to touch and behavioral choice. Cortical responses decrease upon repeated stimulus presentation outside of the behavioral task. Moreover, training improves population encoding of the passage of time, and unexpected deviations in trial timing elicit even stronger responses than touches do. In conclusion, the superficial layers of sensory cortex exhibit a high degree of learning-dependent plasticity and are strongly modulated by non-sensory but behaviorally-relevant features, such as timing and surprise.

Activity in the superficial layers of the sensory cortex is believed to be largely driven by incoming sensory stimuli. Here the authors demonstrate how learning changes neural responses to sensations according to both behavioral relevance and timing, suggesting a high degree of non-sensory modulation.

Efficient Temporal Coding in the Early Visual System: Existing Evidence and Future Directions

While it is universally accepted that the brain makes predictions, there is little agreement about how this is accomplished and under which conditions. Accurate prediction requires neural circuits to learn and store spatiotemporal patterns observed in the natural environment, but it is not obvious how such information should be stored, or encoded. Information theory provides a mathematical formalism that can be used to measure the efficiency and utility of different coding schemes for data transfer and storage. This theory shows that codes become efficient when they remove predictable, redundant spatial and temporal information. Efficient coding has been used to understand retinal computations and may also be relevant to understanding more complicated temporal processing in visual cortex. However, the literature on efficient coding in cortex is varied and can be confusing since the same terms are used to mean different things in different experimental and theoretical contexts. In this work, we attempt to provide a clear summary of the theoretical relationship between efficient coding and temporal prediction, and review evidence that efficient coding principles explain computations in the retina. We then apply the same framework to computations occurring in early visuocortical areas, arguing that data from rodents is largely consistent with the predictions of this model. Finally, we review and respond to criticisms of efficient coding and suggest ways that this theory might be used to design future experiments, with particular focus on understanding the extent to which neural circuits make predictions from efficient representations of environmental statistics.

Visual association cortex links cues with conjunctions of reward and locomotor contexts

Postrhinal cortex (POR) and neighboring lateral visual association areas are necessary for identifying objects and interpreting them in specific contexts, but how POR neurons encode the same object across contexts remains unclear. Here, we imaged excitatory neurons in mouse POR across tens of days prior to and throughout initial cue-reward learning and reversal learning. We assessed responses to the same cue when it was rewarded or unrewarded, during both locomotor and stationary contexts. Surprisingly, a large class of POR neurons were minimally cue-driven prior to learning. After learning, distinct clusters within this class responded selectively to a given cue when presented in a specific conjunction of reward and locomotion contexts. In addition, another class contained clusters of neurons whose cue responses were more transient, insensitive to reward learning, and adapted over thousands of presentations. These two classes of POR neurons may support context-dependent interpretation and context-independent identification of sensory cues.