Predictive processing offers a unifying view of neural computation, proposing that brains continuously anticipate sensory input and update internal models based on prediction errors. In this talk, I will present converging evidence for the computational mechanisms underlying this framework across human neuroscience and deep neural networks. I will begin with recent work showing that large-scale distributed prediction-error encoding in the human brain directly predicts how sensory representations reorganize through predictive learning. I will then turn to PredNet, a popular predictive coding inspired deep network that has been widely used to model real-world biological vision systems. Using dynamic stimuli generated with our Spatiotemporal Style Transfer algorithm, we demonstrate that PredNet relies primarily on low-level spatiotemporal structure and remains insensitive to high-level content, revealing limits in its generalization capacity. Finally, I will discuss new recurrent vision models that integrate top-down feedback connections with intrinsic neural variability, uncovering a dual mechanism for robust sensory coding in which neural variability decorrelates unit responses, while top-down feedback stabilizes network dynamics. Together, these results outline how prediction error signaling and top-down feedback pathways shape adaptive sensory processing in biological and artificial systems.

Zebrafish larva possesses a combination of assets – small dimensions, brain transparency, genetic tractability – which makes it a unique vertebrate model system to probe brain-scale neuronal dynamics. Using light-sheet microscopy, it is currently possible to monitor the activity of the entire brain at cellular resolution using functional calcium imaging, at about 1 full brain/second. The student will harness this unique opportunity to dissect the neural computation at play during sensory-driven navigation. 5-7 days old larvae will be partially restrained in agarose, i.e. with their tail free. Real-time video-monitoring of the tail beats will be used to infer virtual navigational parameters (displacement, reorientation); visual or thermal stimuli will be delivered to the larvae in a manner that will simulate a realistic navigation along light or thermal gradients. During this virtual sensory-driven navigation, the brain activity will be monitored using two-photon light-sheet functional imaging. These experiments will provide rich datasets of whole-brain activity during a complex sensorimotor task. The network dynamics will be analysed in order to extract a finite number of brain states associated with various motor programs. Starting from spontaneous navigation phases (i.e. absence of varying sensory cues), the student will analyse how different sensory cues interfere with the network endogenous dynamics to bias the probability of these different brain states and eventually favor movements along sensory gradients. For more information see: https://www.smartnets-etn.eu/whole-brain-network-dynamics-in-zebrafish-larvae-during-spontaneous-and-sensory-driven-virtual-navigation/

Post-doctoral position in Cognitive Computational Neuroscience at the Adaptive Brain Lab. The role involves combining high field brain imaging (7T fMRI, MR Spectroscopy), electrophysiology (EEG), computational modelling (machine learning, reinforcement learning) and interventions (TMS, tDCS, pharmacology) to understand network dynamics for learning and brain plasticity. The research programme bridges work across scales (local circuits, global networks) and species (humans, rodents) to uncover the neurocomputations that support learning and brain plasticity.

Subsecond timing underlies nearly all sensory and motor activities across species and is critical to survival. While subsecond temporal information has been found across cortical and subcortical regions, it is unclear if it is generated locally and intrinsically or if it is a read out of a centralized clock-like mechanism. Indeed, mechanisms of subsecond timing at the circuit level are largely obscure. Primary sensory areas are well-suited to address these question as they have early access to sensory information and provide minimal processing to it: if temporal information is found in these regions, it is likely to be generated intrinsically and locally. We test this hypothesis by training mice to perform an audio-visual temporal pattern sensory discrimination task as we use 2-photon calcium imaging, a technique capable of recording population level activity at single cell resolution, to record activity in primary visual cortex (V1). We have found significant changes in network dynamics through mice’s learning of the task from naive to middle to expert levels. Changes in network dynamics and behavioral performance are well accounted for by an intrinsic model of timing in which the trajectory of q network through high dimensional state space represents temporal sensory information. Conversely, while we found evidence of other temporal encoding models, such as oscillatory activity, we did not find that they accounted for increased performance but were in fact correlated with the intrinsic model itself. These results provide insight into how subsecond temporal information is encoded mechanistically at the circuit level.

Working memory (WM) is a cognitive function that allows the short-term maintenance and manipulation of information when no longer accessible to the senses. It relies on temporarily storing stimulus features in the activity of neuronal populations. To preserve these dynamics from distraction it has been proposed that pre and post-distraction population activity decomposes into orthogonal subspaces. If orthogonalization is necessary to avoid WM distraction, it should emerge as performance in the task improves. We sought evidence of WM orthogonalization learning and the underlying mechanisms by analyzing calcium imaging data from the prelimbic (PrL) and anterior cingulate (ACC) cortices of mice as they learned to perform an olfactory dual task. The dual task combines an outer Delayed Paired-Association task (DPA) with an inner Go-NoGo task. We examined how neuronal activity reflected the process of protecting the DPA sample information against Go/NoGo distractors. As mice learned the task, we measured the overlap between the neural activity onto the low-dimensional subspaces that encode sample or distractor odors. Early in the training, pre-distraction activity overlapped with both sample and distractor subspaces. Later in the training, pre-distraction activity was strictly confined to the sample subspace, resulting in a more robust sample code. To gain mechanistic insight into how these low-dimensional WM representations evolve with learning we built a recurrent spiking network model of excitatory and inhibitory neurons with low-rank connections. The model links learning to (1) the orthogonalization of sample and distractor WM subspaces and (2) the orthogonalization of each subspace with irrelevant inputs. We validated (1) by measuring the angular distance between the sample and distractor subspaces through learning in the data. Prediction (2) was validated in PrL through the photoinhibition of ACC to PrL inputs, which induced early-training neural dynamics in well-trained animals. In the model, learning drives the network from a double-well attractor toward a more continuous ring attractor regime. We tested signatures for this dynamical evolution in the experimental data by estimating the energy landscape of the dynamics on a one-dimensional ring. In sum, our study defines network dynamics underlying the process of learning to shield WM representations from distracting tasks.

The critical brain hypothesis states that the brain can benefit from operating close to a second-order phase transition. While it has been shown that several computational aspects of sensory information processing (e.g., sensitivity to input) are optimal in this regime, it is still unclear whether these computational benefits of criticality can be leveraged by neural systems performing behaviorally relevant computations. To address this question, we investigate signatures of criticality in networks optimized to perform efficient encoding. We consider a network of leaky integrate-and-fire neurons with synaptic transmission delays and input noise. Previously, it was shown that the performance of such networks varies non-monotonically with the noise amplitude. Interestingly, we find that in the vicinity of the optimal noise level for efficient coding, the network dynamics exhibits signatures of criticality, namely, the distribution of avalanche sizes follows a power law. When the noise amplitude is too low or too high for efficient coding, the network appears either super-critical or sub-critical, respectively. This result suggests that two influential, and previously disparate theories of neural processing optimization—efficient coding, and criticality—may be intimately related

On Thursday, April 27th, we will host Hayriye Cagnan and Philip A. Starr. Hayriye Cagnan, PhD, is an associate professor at the MRC Brain Network Dynamics Unit and University of Oxford. She will tell us about “Therapies orchestrated by patients’ own rhythms”. Philip A. Starr, MD, PhD, is a neurosurgeon and professor of Neurological Surgery at the University of California San Francisco. Besides his scientific presentation on “My evolution in invasive human neurophysiology: from basal ganglia single units to chronic electrocorticography”, he will give us a glimpse at the person behind the science. The talks will be followed by a shared discussion. You can register via talks.stimulatingbrains.org to receive the (free) Zoom link!

Modern electrophysiological recordings simultaneously capture single-unit spiking activities of hundreds of neurons. These neurons exhibit highly complex coordination patterns. Where does this complexity stem from? One candidate is the ubiquitous heterogeneity in connectivity of local neural circuits. Studying neural network dynamics in the linearized regime and using tools from statistical field theory of disordered systems, we derive relations between structure and dynamics that are readily applicable to subsampled recordings of neural circuits: Measuring the statistics of pairwise covariances allows us to infer statistical properties of the underlying connectivity. Applying our results to spontaneous activity of macaque motor cortex, we find that the underlying network operates in a strongly recurrent regime. In this regime, network connectivity is highly heterogeneous, as quantified by a large radius of bulk connectivity eigenvalues. Being close to the point of linear instability, this dynamical regime predicts a rich correlation structure, a large dynamical repertoire, long-range interaction patterns, relatively low dimensionality and a sensitive control of neuronal coordination. These predictions are verified in analyses of spontaneous activity of macaque motor cortex and mouse visual cortex. Finally, we show that even microscopic features of connectivity, such as connection motifs, systematically scale up to determine the global organization of activity in neural circuits.

Electrophysiological recordings during perceptual decision tasks in monkeys suggest that the degree of confidence in a decision is based on a simple neural signal produced by the neural decision process. Attractor neural networks provide an appropriate biophysical modeling framework, and account for the experimental results very well. However, it remains unclear whether attractor neural networks can account for confidence reports in humans. We present the results from an experiment in which participants are asked to perform an orientation discrimination task, followed by a confidence judgment. Here we show that an attractor neural network model quantitatively reproduces, for each participant, the relations between accuracy, response times and confidence. We show that the attractor neural network also accounts for confidence-specific sequential effects observed in the experiment (participants are faster on trials following high confidence trials), as well as non confidence-specific sequential effects. Remarkably, this is obtained as an inevitable outcome of the network dynamics, without any feedback specific to the previous decision (that would result in, e.g., a change in the model parameters before the onset of the next trial). Our results thus suggest that a metacognitive process such as confidence in one’s decision is linked to the intrinsically nonlinear dynamics of the decision-making neural network.

Efficient navigation requires animals to track their position, velocity and heading direction (HD). Some animals’ behavior suggests that they also track uncertainties about these navigational variables, and make strategic use of these uncertainties, in line with a Bayesian computation. Ring-attractor networks have been proposed to estimate and track these navigational variables, for instance in the HD system of the fruit fly Drosophila. However, such networks are not designed to incorporate a notion of uncertainty, and therefore seem unsuited to implement dynamic Bayesian inference. Here, we close this gap by showing that specifically tuned ring-attractor networks can track both a HD estimate and its associated uncertainty, thereby approximating a circular Kalman filter. We identified the network motifs required to integrate angular velocity observations, e.g., through self-initiated turns, and absolute HD observations, e.g., visual landmark inputs, according to their respective reliabilities, and show that these network motifs are present in the connectome of the Drosophila HD system. Specifically, our network encodes uncertainty in the amplitude of a localized bump of neural activity, thereby generalizing standard ring attractor models. In contrast to such standard attractors, however, proper Bayesian inference requires the network dynamics to operate in a regime away from the attractor state. More generally, we show that near-Bayesian integration is inherent in generic ring attractor networks, and that their amplitude dynamics can account for close-to-optimal reliability weighting of external evidence for a wide range of network parameters. This only holds, however, if their connection strengths allow the network to sufficiently deviate from the attractor state. Overall, our work offers a novel interpretation of ring attractor networks as implementing dynamic Bayesian integrators. We further provide a principled theoretical foundation for the suggestion that the Drosophila HD system may implement Bayesian HD tracking via ring attractor dynamics.

Spiking recurrent neural networks (RNNs) are a promising tool for solving a wide variety of complex cognitive and motor tasks, due to their rich temporal dynamics and sparse processing. However training spiking RNNs on dedicated neuromorphic hardware is still an open challenge. This is due mainly to the lack of local, hardware-friendly learning mechanisms that can solve the temporal credit assignment problem and ensure stable network dynamics, even when the weight resolution is limited. These challenges are further accentuated, if one resorts to using memristive devices for in-memory computing to resolve the von-Neumann bottleneck problem, at the expense of a substantial increase in variability in both the computation and the working memory of the spiking RNNs. In this talk, I will present our recent work where we introduced a PyTorch simulation framework of memristive crossbar arrays that enables accurate investigation of such challenges. I will show that recently proposed e-prop learning rule can be used to train spiking RNNs whose weights are emulated in the presented simulation framework. Although e-prop locally approximates the ideal synaptic updates, it is difficult to implement the updates on the memristive substrate due to substantial device non-idealities. I will mention several widely adapted weight update schemes that primarily aim to cope with these device non-idealities and demonstrate that accumulating gradients can enable online and efficient training of spiking RNN on memristive substrates.

Change is ubiquitous in living beings. In particular, the connectome and neural representations can change. Nevertheless behaviors and memories often persist over long times. In a standard model, associative memories are represented by assemblies of strongly interconnected neurons. For faithful storage these assemblies are assumed to consist of the same neurons over time. We propose a contrasting memory model with complete temporal remodeling of assemblies, based on experimentally observed changes of synapses and neural representations. The assemblies drift freely as noisy autonomous network activity or spontaneous synaptic turnover induce neuron exchange. The exchange can be described analytically by reduced, random walk models derived from spiking neural network dynamics or from first principles. The gradual exchange allows activity-dependent and homeostatic plasticity to conserve the representational structure and keep inputs, outputs and assemblies consistent. This leads to persistent memory. Our findings explain recent experimental results on temporal evolution of fear memory representations and suggest that memory systems need to be understood in their completeness as individual parts may constantly change.

Executive control processes and flexible behaviors rely on the integrity of, and dynamic interactions between, large-scale functional brain networks. The right insular cortex is a critical component of a salience/midcingulo-insular network that is thought to mediate interactions between brain networks involved in externally oriented (central executive/lateral frontoparietal network) and internally oriented (default mode/medial frontoparietal network) processes. How these brain systems reconfigure with development is a critical question for cognitive neuroscience, with implications for neurodevelopmental pathologies affecting brain connectivity. I will describe studies examining how brain network dynamics support flexible behaviors in typical and atypical development, presenting evidence suggesting a unique role for the dorsal anterior insular from studies of meta-analytic connectivity modeling, dynamic functional connectivity, and structural connectivity. These findings from adults, typically developing children, and children with autism suggest that structural and functional maturation of insular pathways is a critical component of the process by which human brain networks mature to support complex, flexible cognitive processes throughout the lifespan.

Our group focuses on brain computation, physiology and evolution, with a particular focus on network dynamics, sleep (evolution and mechanistic underpinnings), cortical computation (through the study of ancestral cortices), and sensorimotor processing. This talk will describe our recent results on the remarkable camouflage behavior of cuttlefish (action) and on brain activity in REM and NonREM in lizards (sleep). Both topics will focus on aspects of circuit dynamics.

Our group focuses on brain computation, physiology and evolution, with a particular focus on network dynamics, sleep (evolution and mechanistic underpinnings), cortical computation (through the study of ancestral cortices), and sensorimotor processing. This talk will describe our recent results on the remarkable camouflage behavior of cuttlefish (action) and on brain activity in REM and NonREM in lizards (sleep). Both topics will focus on aspects of circuit dynamics.

Ever since Bliss and Lømo discovered the phenomenon of long-term potentiation (LTP) in rabbit dentate gyrus in the 1960s, Hebb’s rule—neurons that fire together wire together—gained popularity to explain learning and memory. Accumulating evidence, however, suggests that neural activity is homeostatically regulated. Homeostatic mechanisms are mostly interpreted to stabilize network dynamics. However, recent theoretical work has shown that linking the activity of a neuron to its connectivity within the network provides a robust alternative implementation of Hebb’s rule, although entirely based on negative feedback. In this setting, both natural and artificial stimulation of neurons can robustly trigger network rewiring. We used computational models of plastic networks to simulate the complex temporal dynamics of network rewiring in response to external stimuli. In parallel, we performed optogenetic stimulation experiments in the mouse anterior cingulate cortex (ACC) and subsequently analyzed the temporal profile of morphological changes in the stimulated tissue. Our results suggest that the new theoretical framework combining neural activity homeostasis and structural plasticity provides a consistent explanation of our experimental observations.

Neural responses are variable: even under identical experimental conditions, single neuron and population responses typically differ from trial to trial and across time. Recent work has demonstrated that this variability has predictable structure, can be modulated by sensory input and behaviour, and bears critical signatures of the underlying network dynamics and computations. However, current methods for characterising neural variability are primarily geared towards sensory coding in the laboratory: they require trials with repeatable experimental stimuli and behavioural covariates. In addition, they make strong assumptions about the parametric form of variability, rely on assumption-free but data-inefficient histogram-based approaches, or are altogether ill-suited for capturing variability modulation by covariates. Here we present a universal probabilistic spike count model that eliminates these shortcomings. Our method uses scalable Bayesian machine learning techniques to model arbitrary spike count distributions (SCDs) with flexible dependence on observed as well as latent covariates. Without requiring repeatable trials, it can flexibly capture covariate-dependent joint SCDs, and provide interpretable latent causes underlying the statistical dependencies between neurons. We apply the model to recordings from a canonical non-sensory neural population: head direction cells in the mouse. We find that variability in these cells defies a simple parametric relationship with mean spike count as assumed in standard models, its modulation by external covariates can be comparably strong to that of the mean firing rate, and slow low-dimensional latent factors explain away neural correlations. Our approach paves the way to understanding the mechanisms and computations underlying neural variability under naturalistic conditions, beyond the realm of sensory coding with repeatable stimuli.

Dysregulation of emotional processing and its integration with cognitive functions are central features of many mental/emotional disorders associated both with externalizing problems (aggressive, antisocial behaviors) and internalizing problems (anxiety, depression). As Dr. Joseph LeDoux, our invited speaker of this program, wrote in his famous book “Synaptic self: How Our Brains Become Who We Are”—the brain’s synapses—are the channels through which we think, act, imagine, feel, and remember. Synapses encode the essence of personality, enabling each of us to function as a distinctive, integrated individual from moment to moment. Thus, exploring the functioning of synapses leads to the understanding of the mechanism of (patho)physiological function of our brain. In this context, we have investigated the pathophysiology of psychiatric disorders, with particular emphasis on the synaptic function of model mice of various psychiatric disorders such as schizophrenia, autism, depression, and PTSD. Our current interest is how synaptic inputs are integrated to generate the action potential. Because the spatiotemporal organization of neuronal firing is crucial for information processing, but how thousands of inputs to the dendritic spines drive the firing remains a central question in neuroscience. We identified a distinct pattern of synaptic integration in the disease-related models, in which extra-large (XL) spines generate NMDA spikes within these spines, which was sufficient to drive neuronal firing. We experimentally and theoretically observed that XL spines negatively correlated with working memory. Our work offers a whole new concept for dendritic computation and network dynamics, and the understanding of psychiatric research will be greatly reconsidered. The second half of my talk is the development of a novel synaptic tool. Because, no matter how beautifully we can illuminate the spine morphology and how accurately we can quantify the synaptic integration, the links between synapse and brain function remain correlational. In order to challenge the causal relationship between synapse and brain function, we established AS-PaRac1, which is unique not only because it can specifically label and manipulate the recently potentiated dendritic spine (Hayashi-Takagi et al, 2015, Nature). With use of AS-PaRac1, we developed an activity-dependent simultaneous labeling of the presynaptic bouton and the potentiated spines to establish “functional connectomics” in a synaptic resolution. When we apply this new imaging method for PTSD model mice, we identified a completely new functional neural circuit of brain region A→B→C with a very strong S/N in the PTSD model mice. This novel tool of “functional connectomics” and its photo-manipulation could open up new areas of emotional/psychiatric research, and by extension, shed light on the neural networks that determine who we are.

In addition to its role in the learning and expression of conditioned behavior, the amygdala has long been implicated in the regulation of persistent states, such as anxiety and drive. Yet, it is not evident what projections of the neuronal activity capture the functional role of the network across such different timescales, specifically when behavior and neuronal space are complex and high-dimensional. We applied a data-driven dynamical approach for the analysis of calcium imaging data from the basolateral amygdala, collected while mice performed complex, self-paced behaviors, including spatial exploration, free social interaction, and goal directed actions. The seemingly complex network dynamics was effectively described by a hierarchical, modular structure, that corresponded to behavior on multiple timescales. Our results describe the response of the network activity to perturbations along different dimensions and the interplay between slow, state-like representation and the fast processing of specific events and actions schemes. We suggest hierarchical dynamical models offer a unified framework to capture the involvement of the amygdala in transitions between persistent states underlying such different functions as sensory associative learning, action selection and emotional processing. * Work done in collaboration with Jan Gründemann, Sol Fustinana, Alejandro Tsai and Julien Courtin (@theLüthiLab)

TheVirtualBrain (TVB; thevirtualbrain.org) is a software platform for simulating whole-brain network dynamics. TVB models link biophysical parameters at the cellular level with systems-level functional neuroimaging signals. Data available from animal models can provide vital constraints for the linkage across spatial and temporal scales. I will describe the construction of a macaque cortical connectome as an initial step towards a comprehensive multi-scale macaque TVB model. I will also describe our process of validating the connectome and show an example simulation of macaque resting-state dynamics using TVB. This connectome opens the opportunity for the addition of other available data from the macaque, such as electrophysiological recordings and receptor distributions, to inform multi-scale models of brain dynamics. Future work will include extensions to neurological conditions and other nonhuman primate species.

The developmental journey of cortical interneurons encounters several activity-dependent milestones. During the early postnatal period in developing mice, GABAergic neurons are transient preferential recipients of thalamic inputs and undergo activity-dependent migration arrest, wiring and programmed cell-death. But cortical GABAergic neurons are also specified by very early developmental programs. For example, the earliest born GABAergic neurons develop into hub cells coordinating spontaneous activity in hippocampal slices. Despite their importance for the emergence of sensory experience, their role in coordinating network dynamics, and the role of activity in their integration into cortical networks, the collective in vivo dynamics of GABAergic neurons during the neonatal postnatal period remain unknown. Here, I will present data related to the coordinated activity between GABAergic cells of the mouse barrel cortex and hippocampus in non-anesthetized pups using the recent development of all optical methods to record and manipulate neuronal activity in vivo. I will show that the functional structure of developing GABAergic circuits is remarkably patterned, with segregated assemblies of prospective parvalbumin neurons and highly connected hub cells, both shaped by sensory-dependent processes.

Many networks in the brain are sparsely connected, and the brain eliminates synapses during development and learning. How could the brain decide which synapses to prune? In a recurrent network, determining the importance of a synapse between two neurons is a difficult computational problem, depending on the role that both neurons play and on all possible pathways of information flow between them. Noise is ubiquitous in neural systems, and often considered an irritant to be overcome. In the first part of this talk, I will suggest that noise could play a functional role in synaptic pruning, allowing the brain to probe network structure and determine which synapses are redundant. I will introduce a simple, local, unsupervised plasticity rule that either strengthens or prunes synapses using only synaptic weight and the noise-driven covariance of the neighboring neurons. For a subset of linear and rectified-linear networks, this rule provably preserves the spectrum of the original matrix and hence preserves network dynamics even when the fraction of pruned synapses asymptotically approaches 1. The plasticity rule is biologically-plausible and may suggest a new role for noise in neural computation. Time permitting, I will then turn to the problem of extracting structure from neural population data sets using dimensionality reduction methods. I will argue that nonlinear structures naturally arise in neural data and show how these nonlinearities cause linear methods of dimensionality reduction, such as Principal Components Analysis, to fail dramatically in identifying low-dimensional structure.

The description of neural computations currently relies on two competing views: (i) a classical single-cell view that aims to relate the activity of individual neurons to sensory or behavioural variables, and organize them into functional classes; (ii) a more recent population view that instead characterises computations in terms of collective neural trajectories, and focuses on the dimensionality of these trajectories as animals perform tasks. How the two key concepts of functional cell classes and low-dimensional trajectories interact to shape neural computations is however at present not understood. Here I will address this question by combining machine-learning tools for training recurrent neural networks with reverse-engineering and theoretical analyses of network dynamics.