Normalization as a canonical neural computation

Neuroscientific insights about computer vision models: a concise review

The development of biologically-inspired computational models has been the focus of study ever since the artificial neuron was introduced by McCulloch and Pitts in 1943. However, a scrutiny of literature reveals that most attempts to replicate the highly efficient and complex biological visual system have been futile or have met with limited success. The recent state-of the-art computer vision models, such as pre-trained deep neural networks and vision transformers, may not be biologically inspired per se. Nevertheless, certain aspects of biological vision are still found embedded, knowingly or unknowingly, in the architecture and functioning of these models. This paper explores several principles related to visual neuroscience and the biological visual pathway that resonate, in some manner, in the architectural design and functioning of contemporary computer vision models. The findings of this survey can provide useful insights for building futuristic bio-inspired computer vision models. The survey is conducted from a historical perspective, tracing the biological connections of computer vision models starting with the basic artificial neuron to modern technologies such as deep convolutional neural network (CNN) and spiking neural networks (SNN). One spotlight of the survey is a discussion on biologically plausible neural networks and bio-inspired unsupervised learning mechanisms adapted for computer vision tasks in recent times.

Neural pathways and computations that achieve stable contrast processing tuned to natural scenes

Natural scenes are highly dynamic, challenging the reliability of visual processing. Yet, humans and many animals perform accurate visual behaviors, whereas computer vision devices struggle with rapidly changing background luminance. How does animal vision achieve this? Here, we reveal the algorithms and mechanisms of rapid luminance gain control in Drosophila, resulting in stable visual processing. We identify specific transmedullary neurons as the site of luminance gain control, which pass this property to direction-selective cells. The circuitry further involves wide-field neurons, matching computational predictions that local spatial pooling drive optimal contrast processing in natural scenes when light conditions change rapidly. Experiments and theory argue that a spatially pooled luminance signal achieves luminance gain control via divisive normalization. This process relies on shunting inhibition using the glutamate-gated chloride channel GluClα. Our work describes how the fly robustly processes visual information in dynamically changing natural scenes, a common challenge of all visual systems.

Calcium-permeable AMPA receptors govern PV neuron feature selectivity

The brain helps us survive by forming internal representations of the external world1,2. Excitatory cortical neurons are often precisely tuned to specific external stimuli3,4. However, inhibitory neurons, such as parvalbumin-positive (PV) interneurons, are generally less selective5. PV interneurons differ from excitatory neurons in their neurotransmitter receptor subtypes, including AMPA (α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid) receptors (AMPARs)6,7. Excitatory neurons express calcium-impermeable AMPARs that contain the GluA2 subunit (encoded by GRIA2), whereas PV interneurons express receptors that lack the GluA2 subunit and are calcium-permeable (CP-AMPARs). Here we demonstrate a causal relationship between CP-AMPAR expression and the low feature selectivity of PV interneurons. We find low expression stoichiometry of GRIA2 mRNA relative to other subunits in PV interneurons that is conserved across ferrets, rodents, marmosets and humans, and causes abundant CP-AMPAR expression. Replacing CP-AMPARs in PV interneurons with calcium-impermeable AMPARs increased their orientation selectivity in the visual cortex. Manipulations to induce sparse CP-AMPAR expression demonstrated that this increase was cell-autonomous and could occur with changes beyond development. Notably, excitatory-PV interneuron connectivity rates and unitary synaptic strength were unaltered by CP-AMPAR removal, which suggested that the selectivity of PV interneurons can be altered without markedly changing connectivity. In Gria2-knockout mice, in which all AMPARs are calcium-permeable, excitatory neurons showed significantly degraded orientation selectivity, which suggested that CP-AMPARs are sufficient to drive lower selectivity regardless of cell type. Moreover, hippocampal PV interneurons, which usually exhibit low spatial tuning, became more spatially selective after removing CP-AMPARs, which indicated that CP-AMPARs suppress the feature selectivity of PV interneurons independent of modality. These results reveal a new role of CP-AMPARs in maintaining low-selectivity sensory representation in PV interneurons and implicate a conserved molecular mechanism that distinguishes this cell type in the neocortex.

GABAergic inhibition in human hMT+ predicts visuo-spatial intelligence mediated through the frontal cortex

The prevailing opinion emphasizes fronto-parietal network (FPN) is key in mediating general fluid intelligence (gF). Meanwhile, recent studies show that human MT complex (hMT+), located at the occipito-temporal border and involved in 3D perception processing, also plays a key role in gF. However, the underlying mechanism is not clear, yet. To investigate this issue, our study targets visuo-spatial intelligence, which is considered to have high loading on gF. We use ultra-high field magnetic resonance spectroscopy (MRS) to measure GABA/Glu concentrations in hMT+ combining resting-state fMRI functional connectivity (FC), behavioral examinations including hMT+ perception suppression test and gF subtest in visuo-spatial component. Our findings show that both GABA in hMT+ and frontal-hMT+ functional connectivity significantly correlate with the performance of visuo-spatial intelligence. Further, serial mediation model demonstrates that the effect of hMT+ GABA on visuo-spatial gF is fully mediated by the hMT+ frontal FC. Together our findings highlight the importance in integrating sensory and frontal cortices in mediating the visuo-spatial component of general fluid intelligence.

Exploring Nonlinear Dynamics In Brain Functionality Through Phase Portraits And Fuzzy Recurrence Plots

Much of the complexity and diversity found in nature are driven by nonlinear phenomena, and this holds true for the brain. Nonlinear dynamics theory has been successfully utilized in explaining brain functions from a biophysics standpoint, and the field of statistical physics continues to make substantial progress in understanding brain connectivity and function. This study delves into complex brain functional connectivity using biophysical nonlinear dynamics approaches. We aim to uncover hidden information in high-dimensional and nonlinear neural signals, with the hope of providing a useful tool for analyzing information transitions in functionally complex networks. By utilizing phase portraits and fuzzy recurrence plots, we investigated the latent information in the functional connectivity of complex brain networks. Our numerical experiments, which include synthetic linear dynamics neural time series and a biophysically realistic neural mass model, showed that phase portraits and fuzzy recurrence plots are highly sensitive to changes in neural dynamics, and they can also be used to predict functional connectivity based on structural connectivity. Furthermore, the results showed that phase trajectories of neuronal activity encode low-dimensional dynamics, and the geometric properties of the limit-cycle attractor formed by the phase portraits can be used to explain the neurodynamics. Additionally, our results showed that the phase portrait and fuzzy recurrence plots can be used as functional connectivity descriptors, and both metrics were able to capture and explain nonlinear dynamics behavior during specific cognitive tasks. In conclusion, our findings suggest that phase portraits and fuzzy recurrence plots could be highly effective as functional connectivity descriptors, providing valuable insights into nonlinear dynamics in the brain.

Distractor effects in decision making are related to the individual's style of integrating choice attributes

Humans make irrational decisions in the presence of irrelevant distractor options. There is little consensus on whether decision making is facilitated or impaired by the presence of a highly rewarding distractor, or whether the distractor effect operates at the level of options' component attributes rather than at the level of their overall value. To reconcile different claims, we argue that it is important to consider the diversity of people's styles of decision making and whether choice attributes are combined in an additive or multiplicative way. Employing a multi-laboratory dataset investigating the same experimental paradigm, we demonstrated that people used a mix of both approaches and the extent to which approach was used varied across individuals. Critically, we identified that this variability was correlated with the distractor effect during decision making. Individuals who tended to use a multiplicative approach to compute value, showed a positive distractor effect. In contrast, a negative distractor effect (divisive normalisation) was prominent in individuals tending towards an additive approach. Findings suggest that the distractor effect is related to how value is constructed, which in turn may be influenced by task and subject specificities. This concurs with recent behavioural and neuroscience findings that multiple distractor effects co-exist.

Combined manifestation of two geometric visual illusions

The present study continued to investigate whether the effects of length misperception caused by cross-shaped (formed by two pairs of the oppositely oriented Müller-Lyer wings) contextual distractors can be explained by the combined manifestation of two different (i.e., the Müller-Lyer and filled-space) geometric illusions of extent. In psychophysical experiments, the luminance of one pair of wings was randomly changed, while the luminance of the other pair remained constant. Two different distractor orientations were used-when the wings with constant luminance formed the right side of the cross or the left side, otherwise. To separately evaluate the manifestation of the Müller-Lyer illusion under different luminance conditions, two distracting crosses of the same orientation were attached to the lateral stimulus terminators in the first series of experiments. In the following four series, a single distracting cross (with different orientation) was attached to one of the lateral stimulus terminators and various combinations of the constant and background luminance were used. To interpret the experimental data, we used the basic computational principles of previously developed quantitative models of hypothetical visual mechanisms underlying the emergence of the Müller-Lyer illusion and the filled-space illusion. It was shown that the results of theoretical calculations adequately approximate the experimental curves obtained for all modifications of stimuli, which strongly supports the suggestion that the joint manifestations of these two illusions can be considered among the main factors determining the features of the illusion investigated.

The role of binocular disparity and attention in the neural representation of multiple moving stimuli in the visual cortex

Segmenting visual scenes into distinct objects and surfaces is a fundamental visual process, with stereoscopic depth and motion serving as crucial cues. However, how the visual system uses these cues to segment multiple objects is not fully understood. We investigated how neurons in the middle-temporal (MT) cortex of macaque monkeys represent overlapping surfaces at different depths, moving in different directions. Neuronal activity was recorded from three male monkeys during discrimination tasks under varying attention conditions. We found that neuronal responses to overlapping surfaces showed a robust bias toward the binocular disparity of one surface over the other. The disparity bias of a neuron was positively correlated with the neuron's disparity preference for a single surface. In two animals, neurons preferring near disparities of single surfaces (near neurons) showed a near bias for overlapping stimuli, while neurons preferring far disparities (far neurons) showed a far bias. In the third animal, both near and far neurons displayed a near bias, though the near neurons showed a stronger near bias. All three animals exhibited an initial near bias across neurons relative to the average of the responses to the individual surfaces. Although attention modulated neuronal responses, the disparity bias was not caused by attention. We also found that the effect of attention was consistent with object-based, rather than feature-based attention. We proposed a model in which the pool size of the neuron population that weighs the responses to individual stimulus components can be variable. This model is a novel extension of the standard normalization model and provides a unified explanation for the disparity bias across animals. Our results reveal how MT neurons encode multiple stimuli moving at different depths and present new evidence of response modulation by object-based attention. The disparity bias allows subgroups of neurons to preferentially represent individual surfaces of multiple stimuli at different depths, thereby facilitating segmentation.

Exposure to heat wave risks across time and places: Seasonal variations and predictors of feelings of threat across heat wave geographical susceptibility locations

Vulnerability to heat waves and their negative effects on health vary not only due to individual factors but also due to situational factors, such as time and geography. Hence, we explored seasonal variations and predictors of heat wave feelings of threat across different heat wave geographical susceptibility locations in Portugal. A total of 238 Portuguese residents responded to a web-based longitudinal survey: before the summer, during a heat wave in the summer, during the summer, and after the summer. Geographical location was used as an indicator of risk exposure, operationalized as heat wave occurrence susceptibility (low, moderate, high). Heat wave demands and resources perceptions were assessed to compute an indicator of heat wave feelings of threat. During the heat wave, feelings of threat were higher among participants in high-susceptibility locations, with demands outweighing resources perceptions, suggesting greater distress and coping difficulty. Regression analysis suggested that older participants and female participants living in moderate-high-susceptibility locations had greater difficulty in recovering. Heat wave risk perception and positive affect about heat were identified as the most consistent predictors of heat wave feelings of threat, with risk perception increasing and positive affect decreasing such feelings. Participants with (individual and geographical) vulnerability profiles, who had greater difficulty in coping and recovering from heat waves, could benefit from resource-building/enhancing interventions. In a climatic crisis context, monitoring psychological responses to heat waves (e.g., threat) may enable anticipated action to build resilience before, rather than after, the effects become damaging to physical and psychological health.

Syntax-Guided Content-Adaptive Transform for Image Compression

The surge in image data has significantly increased the pressure on storage and transmission, posing new challenges for image compression technology. The structural texture of an image implies its statistical characteristics, which is effective for image encoding and decoding. Consequently, content-adaptive compression methods based on learning can better capture the content attributes of images, thereby enhancing encoding performance. However, learned image compression methods do not comprehensively account for both the global and local correlations among the pixels within an image. Moreover, they are constrained by rate-distortion optimization, which prevents the attainment of a compact representation of image attributes. To address these issues, we propose a syntax-guided content-adaptive transform framework that efficiently captures image attributes and enhances encoding efficiency. Firstly, we propose a syntax-refined side information module that fully leverages syntax and side information to guide the adaptive transformation of image attributes. Moreover, to more thoroughly exploit the global and local correlations in image space, we designed global-local modules, local-global modules, and upsampling/downsampling modules in codecs, further eliminating local and global redundancies. The experimental findings indicate that our proposed syntax-guided content-adaptive image compression model successfully adapts to the diverse complexities of different images, which enhances the efficiency of image compression. Concurrently, the method proposed has demonstrated outstanding performance across three benchmark datasets.

Linking cortical surface area to computational properties in human visual perception

Cortical structure and function are closely linked, shaping the neural basis of human behavior. This study explores how cortical surface area (SA), a structural feature, influences computational properties in human visual perception. Using a combination of psychophysical, neuroimaging, and computational modeling approaches, we find that variations in SA across the parietal and frontal cortices are linked to distinct behavioral patterns in a motion perception task. These differences in behavior correspond to specific parameters within a divisive normalization model, indicating a unique contribution of SA to the spatial organization of cortical circuitry. This work highlights the importance of cortical architecture in modifying computational processes that underlie perception, enhancing our understanding of how structural differences can influence neural function and behavior.

Adaptation in the visual system: Networked fatigue or suppressed prediction error signalling?

Our brains are constantly adapting to changes in our visual environments. Neural adaptation exerts a persistent influence on the activity of sensory neurons and our perceptual experience, however there is a lack of consensus regarding how adaptation is implemented in the visual system. One account describes fatigue-based mechanisms embedded within local networks of stimulus-selective neurons (networked fatigue models). Another depicts adaptation as a product of stimulus expectations (predictive coding models). In this review, I evaluate neuroimaging and psychophysical evidence that poses fundamental problems for predictive coding models of neural adaptation. Specifically, I discuss observations of distinct repetition and expectation effects, as well as incorrect predictions of repulsive adaptation aftereffects made by predictive coding accounts. Based on this evidence, I argue that networked fatigue models provide a more parsimonious account of adaptation effects in the visual system. Although stimulus expectations can be formed based on recent stimulation history, any consequences of these expectations are likely to co-occur (or interact) with effects of fatigue-based adaptation. I conclude by proposing novel, testable hypotheses relating to interactions between fatigue-based adaptation and other predictive processes, focusing on stimulus feature extrapolation phenomena.

Contrast gain control is a reparameterization of a population response curve

Neurons in primary visual cortex (area V1) adapt in different degrees to the average contrast of the environment, suggesting that the representation of visual stimuli may interact with the state of cortical gain control in complex ways. To investigate this possibility, we measured and analyzed the responses of neural populations to visual stimuli as a function of contrast in different environments, each characterized by a unique distribution of contrast. Our findings reveal that, for a given stimulus, the population response can be described by a vector functionr ( g e c ) , where the gaing e is a decreasing function of the mean contrast of the environment. Thus, gain control can be viewed as a reparameterization of a population response curve, which is invariant across environments. Different stimuli are mapped to distinct curves, all originating from a common origin, corresponding to a zero-contrast response. Altogether, our findings provide a straightforward, geometric interpretation of contrast gain control at the population level and show that changes in gain are well coordinated among members of a neural population.

Distinct feedforward and feedback pathways for cell-type specific attention effects

Selective attention is thought to depend on enhanced firing activity in extrastriate areas. Theories suggest that this enhancement depends on selective inter-areal communication via gamma (30-80 Hz) phase-locking. To test this, we simultaneously recorded from different cell types and cortical layers of macaque V1 and V4. We find that while V1-V4 gamma phase-locking between local field potentials increases with attention, the V1 gamma rhythm does not engage V4 excitatory-neurons, but only fast-spiking interneurons in L4 of V4. By contrast, attention enhances V4 spike-rates in both excitatory and inhibitory cells, most strongly in L2/3. The rate increase in L2/3 of V4 precedes V1 in time. These findings suggest enhanced signal transmission with attention does not depend on inter-areal gamma phase-locking and show that the endogenous gamma rhythm has cell-type- and layer-specific effects on downstream target areas. Similar findings were made in the mouse visual system, based on opto-tagging of identified interneurons.

Orthogonality of sensory and contextual categorical dynamics embedded in a continuum of responses from the second somatosensory cortex

How does the brain simultaneously process signals that bring complementary information, like raw sensory signals and their transformed counterparts, without any disruptive interference? Contemporary research underscores the brain's adeptness in using decorrelated responses to reduce such interference. Both neurophysiological findings and artificial neural networks support the notion of orthogonal representation for signal differentiation and parallel processing. Yet, where, and how raw sensory signals are transformed into more abstract representations remains unclear. Using a temporal pattern discrimination task in trained monkeys, we revealed that the second somatosensory cortex (S2) efficiently segregates faithful and transformed neural responses into orthogonal subspaces. Importantly, S2 population encoding for transformed signals, but not for faithful ones, disappeared during a nondemanding version of this task, which suggests that signal transformation and their decoding from downstream areas are only active on-demand. A mechanistic computation model points to gain modulation as a possible biological mechanism for the observed context-dependent computation. Furthermore, individual neural activities that underlie the orthogonal population representations exhibited a continuum of responses, with no well-determined clusters. These findings advocate that the brain, while employing a continuum of heterogeneous neural responses, splits population signals into orthogonal subspaces in a context-dependent fashion to enhance robustness, performance, and improve coding efficiency.

Data-driven modelling of visual receptive fields: comparison between the generalized quadratic model and the nonlinear input model

Objective: Neurons in primary visual cortex (V1) display a range of sensitivity in their response to translations of their preferred visual features within their receptive field: from high specificity to a precise position through to complete invariance. This visual feature selectivity and invariance is frequently modeled by applying a selection of linear spatial filters to the input image, that define the feature selectivity, followed by a nonlinear function that combines the filter outputs, that defines the invariance, to predict the neural response. We compare two such classes of model, that are both popular and parsimonious, the generalized quadratic model (GQM) and the nonlinear input model (NIM). These two classes of model differ primarily in that the NIM can accommodate a greater diversity in the form of nonlinearity that is applied to the outputs of the filters.Approach: We compare the two model types by applying them to data from multielectrode recordings from cat primary visual cortex in response to spatially white Gaussian noise After fitting both classes of model to a database of 342 single units (SUs), we analyze the qualitative and quantitative differences in the visual feature processing performed by the two models and their ability to predict neural response.Main results: We find that the NIM predicts response rates on a held-out data at least as well as the GQM for 95% of SUs. Superior performance occurs predominantly for those units with above average spike rates and is largely due to the NIMs ability to capture aspects of the model's nonlinear function cannot be captured with the GQM rather than differences in the visual features being processed by the two different models.Significance: These results can help guide model choice for data-driven receptive field modelling.

Standard models of spatial vision mispredict edge sensitivity at low spatial frequencies

One well-established characteristic of early visual processing is the contrast sensitivity function (CSF) which describes how sensitivity varies with the spatial frequency (SF) content of the visual input. The CSF prompted the development of a now standard model of spatial vision. It represents the visual input by activity in orientation- and SF selective channels which are nonlinearly recombined to predict a perceptual decision. The standard spatial vision model has been extensively tested with sinusoidal gratings at low contrast because their narrow SF spectra isolate the underlying SF selective mechanisms. It is less studied how well these mechanisms account for sensitivity to more behaviourally relevant stimuli such as sharp edges at high contrast (i.e. object boundaries) which abound in the natural environment and have broader SF spectra. Here, we probe sensitivity to edges (2-AFC, edge localization) in the presence of broadband and narrowband noises. We use Cornsweet luminance profiles with peak frequencies at 0.5, 3 and 9 cpd as edge stimuli. To test how well mechanisms underlying sinusoidal contrast sensitivity can account for edge sensitivity, we implement a single- and a multi-scale model building upon standard spatial vision model components. Both models account for most of the data but also systematically deviate in their predictions, particularly in the presence of pink noise and for the lowest SF edge. These deviations might indicate a transition from contrast- to luminance-based detection at low SFs. Alternatively, they might point to a missing component in current spatial vision models.

The neuron as a direct data-driven controller

In the quest to model neuronal function amid gaps in physiological data, a promising strategy is to develop a normative theory that interprets neuronal physiology as optimizing a computational objective. This study extends current normative models, which primarily optimize prediction, by conceptualizing neurons as optimal feedback controllers. We posit that neurons, especially those beyond early sensory areas, steer their environment toward a specific desired state through their output. This environment comprises both synaptically interlinked neurons and external motor sensory feedback loops, enabling neurons to evaluate the effectiveness of their control via synaptic feedback. To model neurons as biologically feasible controllers which implicitly identify loop dynamics, infer latent states, and optimize control we utilize the contemporary direct data-driven control (DD-DC) framework. Our DD-DC neuron model explains various neurophysiological phenomena: the shift from potentiation to depression in spike-timing-dependent plasticity with its asymmetry, the duration and adaptive nature of feedforward and feedback neuronal filters, the imprecision in spike generation under constant stimulation, and the characteristic operational variability and noise in the brain. Our model presents a significant departure from the traditional, feedforward, instant-response McCulloch-Pitts-Rosenblatt neuron, offering a modern, biologically informed fundamental unit for constructing neural networks.

Signal switching may enhance processing power of the brain

Our ability to perceive multiple objects is mysterious. Sensory neurons are broadly tuned, producing potential overlap in the populations of neurons activated by each object in a scene. This overlap raises questions about how distinct information is retained about each item. We present a novel signal switching theory of neural representation, which posits that neural signals may interleave representations of individual items across time. Evidence for this theory comes from new statistical tools that overcome the limitations inherent to standard time-and-trial-pooled assessments of neural signals. Our theory has implications for diverse domains of neuroscience, including attention, figure binding/scene segregation, oscillations, and divisive normalization. The general concept of switching between functions could also lend explanatory power to theories of grounded cognition.

Valuations of target items are drawn towards unavailable decoy items due to prior expectations

When people make choices, the items they consider are often embedded in a context (of other items). How this context affects the valuation of the specific item is an important question. High-value context might make items appear less attractive because of contrast-the tendency to normalize perception of an object relative to its background-or more attractive because of assimilation-the tendency to group objects together. Alternatively, a high-value context might increase prior expectations about the item's value. Here, we investigated these possibilities. We examined how unavailable context items affect choices between two target items, as well as the willingness-to-pay for single targets. Participants viewed sets of three items for several seconds before the target(s) were highlighted. In both tasks, we found a significant assimilation-like effect where participants were more likely to choose or place a higher value on a target when it was surrounded by higher-value context. However, these context effects were only significant for participants' fastest choices. Using variants of a drift-diffusion model, we established that the unavailable context shifted participants' prior expectations towards the average values of the sets but had an inconclusive effect on their evaluations of the targets during the decision (i.e. drift rates). In summary, we find that people use context to inform their initial valuations. This can improve efficiency by allowing people to get a head start on their decision. However, it also means that the valuation of an item can change depending on the context.

Transcranial Alternating Current Stimulation over Frontal Eye Fields Mimics Attentional Modulation of Visual Processing

Attentional control over sensory processing has been linked to neural alpha oscillations and related inhibition of cerebral cortex. Despite the wide consensus on the functional relevance of alpha oscillations for attention, precise neural mechanisms of how alpha oscillations shape perception and how this top-down modulation is implemented in cortical networks remain unclear. Here, we tested the hypothesis that alpha oscillations in frontal eye fields (FEFs) are causally involved in the top-down regulation of visual processing in humans (male and female). We applied sham-controlled, intermittent transcranial alternating current stimulation (tACS) over bilateral FEF at either 10 Hz (alpha) or 40 Hz (gamma) to manipulate attentional preparation in a visual discrimination task. Under each stimulation condition, we measured psychometric functions for contrast perception and introduced a novel linear mixed modeling approach for statistical control of neurosensory side effects of the electric stimulation. tACS at alpha frequency reduced the slope of the psychometric function, resulting in improved subthreshold and impaired superthreshold contrast perception. Side effects on the psychometric functions were complex and showed large interindividual variability. Controlling for the impact of side effects on the psychometric parameters by using covariates in the linear mixed model analysis reduced this variability and strengthened the perceptual effect. We propose that alpha tACS over FEF mimicked a state of endogenous attention by strengthening a fronto-occipitoparietal network in the alpha band. We speculate that this network modulation enhanced phasic gating in occipitoparietal cortex leading to increased variability of single-trial psychometric thresholds, measurable as a reduction of psychometric slope.

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

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

Origins of noise in both improving and degrading decision making

Noise is a fundamental problem for information processing in neural systems. In decision-making, noise is assumed to have a primary role in errors and stochastic choice behavior. However, little is known about how noise arising from different sources contributes to value coding and choice behaviors, especially when it interacts with neural computation. Here we examine how noise arising early versus late in the choice process differentially impacts context-dependent choice behavior. We found in model simulations that early and late noise predict opposing context effects: under early noise, contextual information enhances choice accuracy; while under late noise, context degrades choice accuracy. Furthermore, we verified these opposing predictions in experimental human choice behavior. Manipulating early and late noise - by inducing uncertainty in option values and controlling time pressure - produced dissociable positive and negative context effects. These findings reconcile controversial experimental findings in the literature reporting either context-driven impairments or improvements in choice performance, suggesting a unified mechanism for context-dependent choice. More broadly, these findings highlight how different sources of noise can interact with neural computations to differentially modulate behavior.

Significance

The current study addresses the role of noise origin in decision-making, reconciling controversies around how decision-making is impacted by context. We demonstrate that different types of noise - either arising early during evaluation or late during option comparison - leads to distinct results: with early noise, context enhances choice accuracy, while with late noise, context impairs it. Understanding these dynamics offers potential strategies for improving decision-making in noisy environments and refining existing neural computation models. Overall, our findings advance our understanding of how neural systems handle noise in essential cognitive tasks, suggest a beneficial role for contextual modulation under certain conditions, and highlight the profound implications of noise structure in decision-making.

Distinct Yet Proximal Face- and Body-Selective Brain Regions Enable Clutter-Tolerant Representations of the Face, Body, and Whole Person

Faces and bodies are processed in separate but adjacent regions in the primate visual cortex. Yet, the functional significance of dividing the whole person into areas dedicated to its face and body components and their neighboring locations remains unknown. Here we hypothesized that this separation and proximity together with a normalization mechanism generate clutter-tolerant representations of the face, body, and whole person when presented in complex multi-category scenes. To test this hypothesis, we conducted a fMRI study, presenting images of a person within a multi-category scene to human male and female participants and assessed the contribution of each component to the response to the scene. Our results revealed a clutter-tolerant representation of the whole person in areas selective for both faces and bodies, typically located at the border between the two category-selective regions. Regions exclusively selective for faces or bodies demonstrated clutter-tolerant representations of their preferred category, corroborating earlier findings. Thus, the adjacent locations of face- and body-selective areas enable a hardwired machinery for decluttering of the whole person, without the need for a dedicated population of person-selective neurons. This distinct yet proximal functional organization of category-selective brain regions enhances the representation of the socially significant whole person, along with its face and body components, within multi-category scenes.

Convolutional neural network models applied to neuronal responses in macaque V1 reveal limited nonlinear processing

Computational models of the primary visual cortex (V1) have suggested that V1 neurons behave like Gabor filters followed by simple nonlinearities. However, recent work employing convolutional neural network (CNN) models has suggested that V1 relies on far more nonlinear computations than previously thought. Specifically, unit responses in an intermediate layer of VGG-19 were found to best predict macaque V1 responses to thousands of natural and synthetic images. Here, we evaluated the hypothesis that the poor performance of lower layer units in VGG-19 might be attributable to their small receptive field size rather than to their lack of complexity per se. We compared VGG-19 with AlexNet, which has much larger receptive fields in its lower layers. Whereas the best-performing layer of VGG-19 occurred after seven nonlinear steps, the first convolutional layer of AlexNet best predicted V1 responses. Although the predictive accuracy of VGG-19 was somewhat better than that of standard AlexNet, we found that a modified version of AlexNet could match the performance of VGG-19 after only a few nonlinear computations. Control analyses revealed that decreasing the size of the input images caused the best-performing layer of VGG-19 to shift to a lower layer, consistent with the hypothesis that the relationship between image size and receptive field size can strongly affect model performance. We conducted additional analyses using a Gabor pyramid model to test for nonlinear contributions of normalization and contrast saturation. Overall, our findings suggest that the feedforward responses of V1 neurons can be well explained by assuming only a few nonlinear processing stages.

How does V1 population activity inform perceptual certainty?

Neural population activity in sensory cortex informs our perceptual interpretation of the environment. Oftentimes, this population activity will support multiple alternative interpretations. The larger the spread of probability over different alternatives, the more uncertain the selected perceptual interpretation. We test the hypothesis that the reliability of perceptual interpretations can be revealed through simple transformations of sensory population activity. We recorded V1 population activity in fixating macaques while presenting oriented stimuli under different levels of nuisance variability and signal strength. We developed a decoding procedure to infer from V1 activity the most likely stimulus orientation as well as the certainty of this estimate. Our analysis shows that response magnitude, response dispersion, and variability in response gain all offer useful proxies for orientation certainty. Of these three metrics, the last one has the strongest association with the decoder's uncertainty estimates. These results clarify that the nature of neural population activity in sensory cortex provides downstream circuits with multiple options to assess the reliability of perceptual interpretations.

Representation and computation in visual working memory

The ability to sustain internal representations of the sensory environment beyond immediate perception is a fundamental requirement of cognitive processing. In recent years, debates regarding the capacity and fidelity of the working memory (WM) system have advanced our understanding of the nature of these representations. In particular, there is growing recognition that WM representations are not merely imperfect copies of a perceived object or event. New experimental tools have revealed that observers possess richer information about the uncertainty in their memories and take advantage of environmental regularities to use limited memory resources optimally. Meanwhile, computational models of visuospatial WM formulated at different levels of implementation have converged on common principles relating capacity to variability and uncertainty. Here we review recent research on human WM from a computational perspective, including the neural mechanisms that support it.

Multimodal gradients of basal forebrain connectivity across the neocortex

The cholinergic innervation of the cortex originates almost entirely from populations of neurons in the basal forebrain (BF). Structurally, the ascending BF cholinergic projections are highly branched, with individual cells targeting multiple different cortical regions. However, it is not known whether the structural organization of basal forebrain projections reflects their functional integration with the cortex. We therefore used high-resolution 7T diffusion and resting state functional MRI in humans to examine multimodal gradients of BF cholinergic connectivity with the cortex. Moving from anteromedial to posterolateral BF, we observed reduced tethering between structural and functional connectivity gradients, with the most pronounced dissimilarity localized in the nucleus basalis of Meynert (NbM). The cortical expression of this structure-function gradient revealed progressively weaker tethering moving from unimodal to transmodal cortex, with the lowest tethering in midcingulo-insular cortex. We used human [ 18 F] fluoroethoxy-benzovesamicol (FEOBV) PET to demonstrate that cortical areas with higher concentrations of cholinergic innervation tend to exhibit lower tethering between BF structural and functional connectivity, suggesting a pattern of increasingly diffuse axonal arborization. Anterograde viral tracing of cholinergic projections and [ 18 F] FEOBV PET in mice confirmed a gradient of axonal arborization across individual BF cholinergic neurons. Like humans, cholinergic neurons with the highest arborization project to cingulo-insular areas of the mouse isocortex. Altogether, our findings reveal that BF cholinergic neurons vary in their branch complexity, with certain subpopulations exhibiting greater modularity and others greater diffusivity in the functional integration of their cortical targets.

Tactile adaptation to orientation produces a robust tilt aftereffect and exhibits crossmodal transfer when tested in vision

Orientation processing is one of the most fundamental functions in both visual and somatosensory perception. Converging findings suggest that orientation processing in both modalities is closely linked: somatosensory neurons share a similar orientation organisation as visual neurons, and the visual cortex has been found to be heavily involved in tactile orientation perception. Hence, we hypothesized that somatosensation would exhibit a similar orientation adaptation effect, and this adaptation effect would be transferable between the two modalities, considering the above-mentioned connection. The tilt aftereffect (TAE) is a demonstration of orientation adaptation and is used widely in behavioural experiments to investigate orientation mechanisms in vision. By testing the classic TAE paradigm in both tactile and crossmodal orientation tasks between vision and touch, we were able to show that tactile perception of orientation shows a very robust TAE, similar to its visual counterpart. We further show that orientation adaptation in touch transfers to produce a TAE when tested in vision, but not vice versa. Additionally, when examining the test sequence following adaptation for serial effects, we observed another asymmetry between the two conditions where the visual test sequence displayed a repulsive intramodal serial dependence effect while the tactile test sequence exhibited an attractive serial dependence. These findings provide concrete evidence that vision and touch engage a similar orientation processing mechanism. However, the asymmetry in the crossmodal transfer of TAE and serial dependence points to a non-reciprocal connection between the two modalities, providing further insights into the underlying processing mechanism.

A dynamic neural resource model bridges sensory and working memory

Probing memory of a complex visual image within a few hundred milliseconds after its disappearance reveals significantly greater fidelity of recall than if the probe is delayed by as little as a second. Classically interpreted, the former taps into a detailed but rapidly decaying visual sensory or 'iconic' memory (IM), while the latter relies on capacity-limited but comparatively stable visual working memory (VWM). While iconic decay and VWM capacity have been extensively studied independently, currently no single framework quantitatively accounts for the dynamics of memory fidelity over these time scales. Here, we extend a stationary neural population model of VWM with a temporal dimension, incorporating rapid sensory-driven accumulation of activity encoding each visual feature in memory, and a slower accumulation of internal error that causes memorized features to randomly drift over time. Instead of facilitating read-out from an independent sensory store, an early cue benefits recall by lifting the effective limit on VWM signal strength imposed when multiple items compete for representation, allowing memory for the cued item to be supplemented with information from the decaying sensory trace. Empirical measurements of human recall dynamics validate these predictions while excluding alternative model architectures. A key conclusion is that differences in capacity classically thought to distinguish IM and VWM are in fact contingent upon a single resource-limited WM store.

Foraging in a non-foraging task: Fitness maximization explains human risk preference dynamics under changing environment

Changes in risk preference have been reported when making a series of independent risky choices or non-foraging economic decisions. Behavioral economics has put forward various explanations for specific changes in risk preference in non-foraging tasks, but a consensus regarding the general principle underlying these effects has not been reached. In contrast, recent studies have investigated human economic risky choices using tasks adapted from foraging theory, which require consideration of past choices and future opportunities to make optimal decisions. In these foraging tasks, human economic risky choices are explained by the ethological principle of fitness maximization, which naturally leads to dynamic risk preference. Here, we conducted two online experiments to investigate whether the principle of fitness maximization can explain risk preference dynamics in a non-foraging task. Participants were asked to make a series of independent risky economic decisions while the environmental richness changed. We found that participants' risk preferences were influenced by the current and past environments, making them more risk-averse during and after the rich environment compared to the poor environment. These changes in risk preference align with fitness maximization. Our findings suggest that the ethological principle of fitness maximization might serve as a generalizable principle for explaining dynamic preferences, including risk preference, in human economic decision-making.

Diverse task-driven modeling of macaque V4 reveals functional specialization towards semantic tasks

Responses to natural stimuli in area V4-a mid-level area of the visual ventral stream-are well predicted by features from convolutional neural networks (CNNs) trained on image classification. This result has been taken as evidence for the functional role of V4 in object classification. However, we currently do not know if and to what extent V4 plays a role in solving other computational objectives. Here, we investigated normative accounts of V4 (and V1 for comparison) by predicting macaque single-neuron responses to natural images from the representations extracted by 23 CNNs trained on different computer vision tasks including semantic, geometric, 2D, and 3D types of tasks. We found that V4 was best predicted by semantic classification features and exhibited high task selectivity, while the choice of task was less consequential to V1 performance. Consistent with traditional characterizations of V4 function that show its high-dimensional tuning to various 2D and 3D stimulus directions, we found that diverse non-semantic tasks explained aspects of V4 function that are not captured by individual semantic tasks. Nevertheless, jointly considering the features of a pair of semantic classification tasks was sufficient to yield one of our top V4 models, solidifying V4's main functional role in semantic processing and suggesting that V4's selectivity to 2D or 3D stimulus properties found by electrophysiologists can result from semantic functional goals.

Temporal dynamics of short-term neural adaptation across human visual cortex

Neural responses in visual cortex adapt to prolonged and repeated stimuli. While adaptation occurs across the visual cortex, it is unclear how adaptation patterns and computational mechanisms differ across the visual hierarchy. Here we characterize two signatures of short-term neural adaptation in time-varying intracranial electroencephalography (iEEG) data collected while participants viewed naturalistic image categories varying in duration and repetition interval. Ventral- and lateral-occipitotemporal cortex exhibit slower and prolonged adaptation to single stimuli and slower recovery from adaptation to repeated stimuli compared to V1-V3. For category-selective electrodes, recovery from adaptation is slower for preferred than non-preferred stimuli. To model neural adaptation we augment our delayed divisive normalization (DN) model by scaling the input strength as a function of stimulus category, enabling the model to accurately predict neural responses across multiple image categories. The model fits suggest that differences in adaptation patterns arise from slower normalization dynamics in higher visual areas interacting with differences in input strength resulting from category selectivity. Our results reveal systematic differences in temporal adaptation of neural population responses between lower and higher visual brain areas and show that a single computational model of history-dependent normalization dynamics, fit with area-specific parameters, accounts for these differences.

In primary visual cortex fMRI responses to chromatic and achromatic stimuli are interdependent and predict contrast detection thresholds

Chromatic and achromatic signals in primary visual cortex have historically been considered independent of each other but have since shown evidence of interdependence. Here, we investigated the combination of two components of a stimulus; an achromatic dynamically changing check background and a chromatic (L-M or S cone) target grating. We found that combinations of chromatic and achromatic signals in primary visual cortex were interdependent, with the dynamic range of responses to chromatic contrast decreasing as achromatic contrast increased. A contrast detection threshold study also revealed interdependence of background and target, with increasing chromatic contrast detection thresholds as achromatic background contrast increased. A model that incorporated a normalising effect of achromatic contrast on chromatic responses, but not vice versa, best predicted our V1 data as well as behavioural thresholds. Further along the visual hierarchy, the dynamic range of chromatic responses was maintained when compared to achromatic responses, which became increasingly compressive.

Perceptual similarity judgments do not predict the distribution of errors in working memory

Population coding models provide a quantitative account of visual working memory (VWM) retrieval errors with a plausible link to the response characteristics of sensory neurons. Recent work has provided an important new perspective linking population coding to variables of signal detection, including d-prime, and put forward a new hypothesis: that the distribution of recall errors on, for example, a color wheel, is a consequence of the psychological similarity between points in that stimulus space, such that the exponential-like psychophysical distance scaling function can fulfil the role of population tuning and obviate the need to fit a tuning width parameter to recall data. Using four different visual feature spaces, we measured psychophysical similarity and memory errors in the same participants. Our results revealed strong evidence for a common source of variability affecting similarity judgments and recall estimates but did not support any consistent relationship between psychophysical similarity functions and VWM errors. At the group level, the responsiveness functions obtained from the psychophysical similarity task diverged strongly from those that provided the best fit to working memory errors. At the individual level, we found convincing evidence against an association between observed and best-fitting similarity functions. Finally, our results show that the newly proposed exponential-like responsiveness function has in general no advantage over the canonical von Mises (circular normal) function assumed by previous population coding models. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Response sub-additivity and variability quenching in visual cortex

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

The mediodorsal thalamus in executive control

Executive control, the ability to organize thoughts and action plans in real time, is a defining feature of higher cognition. Classical theories have emphasized cortical contributions to this process, but recent studies have reinvigorated interest in the role of the thalamus. Although it is well established that local thalamic damage diminishes cognitive capacity, such observations have been difficult to inform functional models. Recent progress in experimental techniques is beginning to enrich our understanding of the anatomical, physiological, and computational substrates underlying thalamic engagement in executive control. In this review, we discuss this progress and particularly focus on the mediodorsal thalamus, which regulates the activity within and across frontal cortical areas. We end with a synthesis that highlights frontal thalamocortical interactions in cognitive computations and discusses its functional implications in normal and pathological conditions.

Unsupervised learning of perceptual feature combinations

In many situations it is behaviorally relevant for an animal to respond to co-occurrences of perceptual, possibly polymodal features, while these features alone may have no importance. Thus, it is crucial for animals to learn such feature combinations in spite of the fact that they may occur with variable intensity and occurrence frequency. Here, we present a novel unsupervised learning mechanism that is largely independent of these contingencies and allows neurons in a network to achieve specificity for different feature combinations. This is achieved by a novel correlation-based (Hebbian) learning rule, which allows for linear weight growth and which is combined with a mechanism for gradually reducing the learning rate as soon as the neuron's response becomes feature combination specific. In a set of control experiments, we show that other existing advanced learning rules cannot satisfactorily form ordered multi-feature representations. In addition, we show that networks, which use this type of learning always stabilize and converge to subsets of neurons with different feature-combination specificity. Neurons with this property may, thus, serve as an initial stage for the processing of ecologically relevant real world situations for an animal.

Active reinforcement learning versus action bias and hysteresis: control with a mixture of experts and nonexperts

Active reinforcement learning enables dynamic prediction and control, where one should not only maximize rewards but also minimize costs such as of inference, decisions, actions, and time. For an embodied agent such as a human, decisions are also shaped by physical aspects of actions. Beyond the effects of reward outcomes on learning processes, to what extent can modeling of behavior in a reinforcement-learning task be complicated by other sources of variance in sequential action choices? What of the effects of action bias (for actions per se) and action hysteresis determined by the history of actions chosen previously? The present study addressed these questions with incremental assembly of models for the sequential choice data from a task with hierarchical structure for additional complexity in learning. With systematic comparison and falsification of computational models, human choices were tested for signatures of parallel modules representing not only an enhanced form of generalized reinforcement learning but also action bias and hysteresis. We found evidence for substantial differences in bias and hysteresis across participants-even comparable in magnitude to the individual differences in learning. Individuals who did not learn well revealed the greatest biases, but those who did learn accurately were also significantly biased. The direction of hysteresis varied among individuals as repetition or, more commonly, alternation biases persisting from multiple previous actions. Considering that these actions were button presses with trivial motor demands, the idiosyncratic forces biasing sequences of action choices were robust enough to suggest ubiquity across individuals and across tasks requiring various actions. In light of how bias and hysteresis function as a heuristic for efficient control that adapts to uncertainty or low motivation by minimizing the cost of effort, these phenomena broaden the consilient theory of a mixture of experts to encompass a mixture of expert and nonexpert controllers of behavior.

No position-specific interference from prior lists in cued recognition: A challenge for position coding (and other) theories of serial memory

Position-specific intrusions of items from prior lists are rare but important phenomena that distinguish broad classes of theory in serial memory. They are uniquely predicted by position coding theories, which assume items on all lists are associated with the same set of codes representing their positions. Activating a position code activates items associated with it in current and prior lists in proportion to their distance from the activated position. Thus, prior list intrusions are most likely to come from the coded position. Alternative "item dependent" theories based on associations between items and contexts built from items have difficulty accounting for the position specificity of prior list intrusions. We tested the position coding account with a position-cued recognition task designed to produce prior list interference. Cuing a position should activate a position code, which should activate items in nearby positions in the current and prior lists. We presented lures from the prior list to test for position-specific activation in response time and error rate; lures from nearby positions should interfere more. We found no evidence for such interference in 10 experiments, falsifying the position coding prediction. We ran two serial recall experiments with the same materials and found position-specific prior list intrusions. These results challenge all theories of serial memory: Position coding theories can explain the prior list intrusions in serial recall and but not the absence of prior list interference in cued recognition. Item dependent theories can explain the absence of prior list interference in cued recognition but cannot explain the occurrence of prior list intrusions in serial recall.

Matching provides efficient decisions

How humans and animals distribute their behavior across choice options has been of key interest to economics, psychology, ecology, and related fields. Neoclassical and behavioral economics have provided prescriptions for how decision-makers can maximize their reward or utility, but these formalisms are used by decision-makers rarely. Instead, individuals allocate their behavior in proportion to the worth of their options, a phenomenon captured by the generalized matching law. Why biological decision-makers adopt this strategy has been unclear. To provide insight into this issue, this article evaluates the performance of matching across a broad spectrum of decision situations, using simulations. Matching is found to attain a high or near-optimal gain, and the strategy achieves this level of performance following a single evaluation of the decision options. Thus, matching provides highly efficient decisions across a wide range of choice environments. This result offers a quantitative explanation for the broad adoption of matching by biological decision-makers.

Matching provides efficient decisions

How humans and animals distribute their behavior across choice options has been of key interest to economics, psychology, ecology, and related fields. Neoclassical and behavioral economics have provided prescriptions for how decision-makers can maximize their reward or utility, but these formalisms are used by decision-makers rarely. Instead, individuals allocate their behavior in proportion to the worth of their options, a phenomenon captured by the generalized matching law. Why biological decision-makers adopt this strategy has been unclear. To provide insight into this issue, this article evaluates the performance of matching across a broad spectrum of decision situations, using simulations. Matching is found to attain a high or near-optimal gain, and the strategy achieves this level of performance following a single evaluation of the decision options. Thus, matching provides highly efficient decisions across a wide range of choice environments. This result offers a quantitative explanation for the broad adoption of matching by biological decision-makers.

Investigating cross-orientation inhibition with continuous tracking

We investigated cross-orientation inhibition with the recently developed continuous tracking technique. We designed an experiment where participants tracked the horizontal motion of a narrow vertical grating. The target was superimposed on one of three different backgrounds, in separate sessions: a uniform gray background or a sinusoidal grating oriented either parallel or orthogonal to the target. Both mask and target where phase reversed. We cross-correlated target and mouse movements and compared the peaks and lags of response with the different masks. Our results are in agreement with previous findings on cross-orientation inhibition: The orthogonal mask had a weak effect on the peaks and lags of correlation as a function of target contrast, consistently with a divisive effect of the mask, while the parallel mask acted subtractively on the response. Interestingly, lags of correlation decreased approximately linearly with contrast, with decrements of the order of 100 ms, even at 10 times the detection threshold, confirming that it is possible to investigate behavioral differences above threshold using the continuous tracking paradigm.

Distinguishing examples while building concepts in hippocampal and artificial networks

The hippocampal subfield CA3 is thought to function as an auto-associative network that stores experiences as memories. Information from these experiences arrives directly from the entorhinal cortex as well as indirectly through the dentate gyrus, which performs sparsification and decorrelation. The computational purpose for these dual input pathways has not been firmly established. We model CA3 as a Hopfield-like network that stores both dense, correlated encodings and sparse, decorrelated encodings. As more memories are stored, the former merge along shared features while the latter remain distinct. We verify our model's prediction in rat CA3 place cells, which exhibit more distinct tuning during theta phases with sparser activity. Finally, we find that neural networks trained in multitask learning benefit from a loss term that promotes both correlated and decorrelated representations. Thus, the complementary encodings we have found in CA3 can provide broad computational advantages for solving complex tasks.

Decomposing Neural Circuit Function into Information Processing Primitives

It is challenging to measure how specific aspects of coordinated neural dynamics translate into operations of information processing and, ultimately, cognitive functions. An obstacle is that simple circuit mechanisms-such as self-sustained or propagating activity and nonlinear summation of inputs-do not directly give rise to high-level functions. Nevertheless, they already implement simple the information carried by neural activity. Here, we propose that distinct functions, such as stimulus representation, working memory, or selective attention, stem from different combinations and types of low-level manipulations of information or information processing primitives. To test this hypothesis, we combine approaches from information theory with simulations of multi-scale neural circuits involving interacting brain regions that emulate well-defined cognitive functions. Specifically, we track the information dynamics emergent from patterns of neural dynamics, using quantitative metrics to detect where and when information is actively buffered, transferred or nonlinearly merged, as possible modes of low-level processing (storage, transfer and modification). We find that neuronal subsets maintaining representations in working memory or performing attentional gain modulation are signaled by their boosted involvement in operations of information storage or modification, respectively. Thus, information dynamic metrics, beyond detecting which network units participate in cognitive processing, also promise to specify how and when they do it, that is, through which type of primitive computation, a capability that may be exploited for the analysis of experimental recordings.

Computational model links normalization to chemoarchitecture in the human visual system

A goal of cognitive neuroscience is to provide computational accounts of brain function. Canonical computations-mathematical operations used by the brain in many contexts-fulfill broad information-processing needs by varying their algorithmic parameters. A key question concerns the identification of biological substrates for these computations and their algorithms. Chemoarchitecture-the spatial distribution of neurotransmitter receptor densities-shapes brain function. Here, we propose that local variations in specific receptor densities implement algorithmic modulations of canonical computations. To test this hypothesis, we combine mathematical modeling of brain responses with chemoarchitecture data. We compare parameters of divisive normalization obtained from 7-tesla functional magnetic resonance imaging with receptor density maps obtained from positron emission tomography. We find evidence that serotonin and γ-aminobutyric acid receptor densities are the biological substrate for algorithmic modulations of divisive normalization in the human visual system. Our model links computational and biological levels of vision, explaining how canonical computations allow the brain to fulfill broad information-processing needs.

Decentralized Neural Circuits of Multisensory Information Integration in the Brain

The brain combines multisensory inputs together to obtain a complete and reliable description of the world. Recent experiments suggest that several interconnected multisensory brain areas are simultaneously involved to integrate multisensory information. It was unknown how these mutually connected multisensory areas achieve multisensory integration. To answer this question, using biologically plausible neural circuit models we developed a decentralized system for information integration that comprises multiple interconnected multisensory brain areas. Through studying an example of integrating visual and vestibular cues to infer heading direction, we show that such a decentralized system is well consistent with experimental observations. In particular, we demonstrate that this decentralized system can optimally integrate information by implementing sampling-based Bayesian inference. The Poisson variability of spike generation provides appropriate variability to drive sampling, and the interconnections between multisensory areas store the correlation prior between multisensory stimuli. The decentralized system predicts that optimally integrated information emerges locally from the dynamics of the communication between brain areas and sheds new light on the interpretation of the connectivity between multisensory brain areas.

Divisively normalized neuronal processing of uncertain visual feedback for visuomotor learning

When encountering a visual error during a reaching movement, the motor system improves the motor command for the subsequent trial. This improvement is impaired by visual error uncertainty, which is considered evidence that the motor system optimally estimates the error. However, how such statistical computation is accomplished remains unclear. Here, we propose an alternative scheme implemented with a divisive normalization (DN): the responses of neuronal elements are normalized by the summed activity of the population. This scheme assumes that when an uncertain visual error is provided by multiple cursors, the motor system processes the error conveyed by each cursor and integrates the information using DN. The DN model reproduced the patterns of learning response to 1-3 cursor errors and the impairment of learning response with visual error uncertainty. This study provides a new perspective on how the motor system updates motor commands according to uncertain visual error information.

Normalization in mouse primary visual cortex

When multiple stimuli appear together in the receptive field of a visual cortical neuron, the response is typically close to the average of that neuron's response to each individual stimulus. The departure from a linear sum of each individual response is referred to as normalization. In mammals, normalization has been best characterized in the visual cortex of macaques and cats. Here we study visually evoked normalization in the visual cortex of awake mice using imaging of calcium indicators in large populations of layer 2/3 (L2/3) V1 excitatory neurons and electrophysiological recordings across layers in V1. Regardless of recording method, mouse visual cortical neurons exhibit normalization to varying degrees. The distributions of normalization strength are similar to those described in cats and macaques, albeit slightly weaker on average.

A power law describes the magnitude of adaptation in neural populations of primary visual cortex

How do neural populations adapt to the time-varying statistics of sensory input? We used two-photon imaging to measure the activity of neurons in mouse primary visual cortex adapted to different sensory environments, each defined by a distinct probability distribution over a stimulus set. We find that two properties of adaptation capture how the population response to a given stimulus, viewed as a vector, changes across environments. First, the ratio between the response magnitudes is a power law of the ratio between the stimulus probabilities. Second, the response direction to a stimulus is largely invariant. These rules could be used to predict how cortical populations adapt to novel, sensory environments. Finally, we show how the power law enables the cortex to preferentially signal unexpected stimuli and to adjust the metabolic cost of its sensory representation to the entropy of the environment.