Search

Hippocampal CA3 is central to memory formation and retrieval. Although various network mechanisms have been proposed, direct evidence is lacking. Using intracellular V m recordings and optogenetic manipulations in behaving mice, we found that CA3 place-field activity is produced by a symmetric form of behavioral timescale synaptic plasticity (BTSP) at recurrent synapses among CA3 pyramidal neurons but not at synapses from the dentate gyrus (DG). Additional manipulations revealed that excitatory input from the entorhinal cortex (EC) but not the DG was required to update place cell activity based on the animal's movement. These data were captured by a computational model that used BTSP and an external updating input to produce attractor dynamics under online learning conditions. Theoretical analyses further highlight the superior memory storage capacity of such networks, especially when dealing with correlated input patterns. This evidence elucidates the cellular and circuit mechanisms of learning and memory formation in the hippocampus. [Display omitted] • CA3 intracellular V m recordings and optogenetic manipulations in behaving mice • Symmetric BTSP at CA3-CA3 recurrent synapses builds attractor dynamics within CA3 • External EC inputs update the CA3 attractor dynamics in accordance with behavior • A neuronal network built by BTSP possesses superior memory storage capacity Evidence from behaving mice points to cellular and circuit mechanisms that underlie observed attractor dynamics in hippocampal area CA3. [ABSTRACT FROM AUTHOR]

Internal states drive survival behaviors, but their neural implementation is poorly understood. Recently, we identified a line attractor in the ventromedial hypothalamus (VMH) that represents a state of aggressiveness. Line attractors can be implemented by recurrent connectivity or neuromodulatory signaling, but evidence for the latter is scant. Here, we demonstrate that neuropeptidergic signaling is necessary for line attractor dynamics in this system by using cell-type-specific CRISPR-Cas9-based gene editing combined with single-cell calcium imaging. Co-disruption of receptors for oxytocin and vasopressin in adult VMH Esr1+ neurons that control aggression diminished attack, reduced persistent neural activity, and eliminated line attractor dynamics while only slightly reducing overall neural activity and sex- or behavior-specific tuning. These data identify a requisite role for neuropeptidergic signaling in implementing a behaviorally relevant line attractor in mammals. Our approach should facilitate mechanistic studies in neuroscience that bridge different levels of biological function and abstraction. [Display omitted] • "CRISPRoscopy" integrates calcium imaging and gene editing in the same neuronal cell type • Co-editing of Oxtr and AVPr1a genes in VMHEsr1 neurons reduces aggression but not mounting • A line attractor encoding a scalable, persistent aggressive internal state is eliminated • Line attractor implementation in the hypothalamus requires neuropeptide signaling Oxytocin and vasopressin receptors are required in the ventromedial hypothalamus for persistent neural activity and line attractor dynamics that encode an aggressive internal state. [ABSTRACT FROM AUTHOR]

Networks of interconnected neurons communicating through spiking signals offer the bedrock of neural computations. Our brain's spiking neural networks have the computational capacity to achieve complex pattern recognition and cognitive functions effortlessly. However, solving real-world problems with artificial spiking neural networks (SNNs) has proved to be difficult for a variety of reasons. Crucially, scaling SNNs to large networks and processing large-scale real-world datasets have been challenging, especially when compared to their non-spiking deep learning counterparts. The critical operation that is needed of SNNs is the ability to learn distributed representations from data and use these representations for perceptual, cognitive and memory operations. In this work, we introduce a novel SNN that performs unsupervised representation learning and associative memory operations leveraging Hebbian synaptic and activity- dependent structural plasticity coupled with neuron-units modelled as Poisson spike generators with sparse firing (~1Hz mean and ~100 Hz maximum firing rate). Crucially, the architecture of our model derives from the neocortical columnar organization and combines feed forward projections for learning hidden representations and recurrent projections for forming associative memories. We evaluated the model on properties relevant for attractor-based associative memories such as pattern completion, perceptual rivalry, distortion resistance, and prototype extraction. [ABSTRACT FROM AUTHOR]

A long-standing research goal is to develop computing technologies that mimic the brain's capabilities by implementing computation in electronic systems directly inspired by its structure, function, and operational mechanisms, using low-power, spike-based neural networks. The Loihi neuromorphic processor provides a low-power, large-scale network of programmable silicon neurons for brain-inspired artificial intelligence applications. This paper exploits the Loihi processors and a theory-guided methodology to enable unsupervised learning of spike patterns. Our method ensures efficient and rapid selection of the network's hyperparameters, enabling the neuromorphic processor to generate attractor states through real-time unsupervised learning. Precisely, we follow a fast design process in which we fine-tune network parameters using mean-field theory. Moreover, we measure the network's learning ability regarding its error correction and pattern completion aptitude. Finally, we observe the dynamic energy consumption of the neuron cores for each millisecond of simulation equal to 23 μ J/time step during the learning and recall phase for four attractors composed of 512 excitatory neurons and 256 shared inhibitory neurons. This study showcases how large-scale, low-power digital neuromorphic processors can be quickly programmed to enable the autonomous generation of attractor states. These attractors are fundamental computational primitives that theoretical analysis and experimental evidence indicate as versatile and reusable components suitable for a wide range of cognitive tasks. [ABSTRACT FROM AUTHOR]

Resilience has traditionally been conceptualized as resisting, bouncing back from, and growing from a stressor. However, recent literature has pointed out that these are different processes with bouncing back coming closest to the literal meaning of the term resilience. To detect whether an individual demonstrates one of these three stressor-responses, different analysis strategies have been suggested. However, deeper theoretical explanations for how patterns of resistance, resilience, and growth come about, have been lacking. To address this gap, this paper proposes a coherent framework based on a dynamical systems approach. We first discuss how adapting to stressors emerges from complex interactions between multiple levels of organization within the system. These interactions unfold on different time scales: What appears as resistance on slower or macro scales may actually consist of bouncing back at micro scales that change much faster. Next, we discuss how the different trajectories that distinguish resistance, resilience, and growth can be understood through attractor dynamics. We address the fixed-point attractors, which are commonly used in the resilience literature to detect early warning signals of bifurcations following resilience losses. Moreover, we describe the implications of limit cycles and strange attractors which capture multiple pathways to adapt to stressors that can lead to growth patterns. We conclude that resisting, bouncing back from, or growing from a stressor represent distinct phenomena that can be distinguished both empirically and theoretically from a dynamical systems perspective. These distinctions may drive future development of theoretical models, empirical measurements, and theory-driven interventions. [ABSTRACT FROM AUTHOR]

Errors in stimulus estimation reveal how stimulus representation changes during cognitive processes. Repulsive bias and minimum variance observed near cardinal axes are well-known error patterns typically associated with visual orientation perception. Recent experiments suggest that these errors continuously evolve during working memory, posing a challenge that neither static sensory models nor traditional memory models can address. Here, we demonstrate that these evolving errors, maintaining characteristic shapes, require network interaction between two distinct modules. Each module fulfills efficient sensory encoding and memory maintenance, which cannot be achieved simultaneously in a single-module network. The sensory module exhibits heterogeneous tuning with strong inhibitory modulation reflecting natural orientation statistics. While the memory module, operating alone, supports homogeneous representation via continuous attractor dynamics, the fully connected network forms discrete attractors with moderate drift speed and nonuniform diffusion processes. Together, our work underscores the significance of sensory-memory interaction in continuously shaping stimulus representation during working memory.

Networks of interconnected neurons communicating through spiking signals offer the bedrock of neural computations. Our brain’s spiking neural networks have the computational capacity to achieve complex pattern recognition and cognitive functions effortlessly. However, solving real-world problems with artificial spiking neural networks (SNNs) has proved to be difficult for a variety of reasons. Crucially, scaling SNNs to large networks and processing large-scale real-world datasets have been challenging, especially when compared to their non-spiking deep learning counterparts. The critical operation that is needed of SNNs is the ability to learn distributed representations from data and use these representations for perceptual, cognitive and memory operations. In this work, we introduce a novel SNN that performs unsupervised representation learning and associative memory operations leveraging Hebbian synaptic and activity-dependent structural plasticity coupled with neuron-units modelled as Poisson spike generators with sparse firing (~1 Hz mean and ~100 Hz maximum firing rate). Crucially, the architecture of our model derives from the neocortical columnar organization and combines feedforward projections for learning hidden representations and recurrent projections for forming associative memories. We evaluated the model on properties relevant for attractor-based associative memories such as pattern completion, perceptual rivalry, distortion resistance, and prototype extraction.

Computational analysis of cellular activity has developed largely independently of modern transcriptomic cell typology, but integrating these approaches may be essential for full insight into cellular-level mechanisms underlying brain function and dysfunction. Applying this approach to the habenula (a structure with diverse, intermingled molecular, anatomical, and computational features), we identified encoding of reward-predictive cues and reward outcomes in distinct genetically defined neural populations, including TH+ cells and Tac1+ cells. Data from genetically targeted recordings were used to train an optimized nonlinear dynamical systems model and revealed activity dynamics consistent with a line attractor. High-density, cell-type-specific electrophysiological recordings and optogenetic perturbation provided supporting evidence for this model. Reverse-engineering predicted how Tac1+ cells might integrate reward history, which was complemented by in vivo experimentation. This integrated approach describes a process by which data-driven computational models of population activity can generate and frame actionable hypotheses for cell-type-specific investigation in biological systems.

Everyday life is composed of events organized by changes in contexts, with each event containing an unfolding sequence of occurrences. A major challenge facing our memory systems is how to integrate sequential occurrences within events while also maintaining their details and avoiding over-integration across different contexts. We asked if and how distinct hippocampal subfields come to hierarchically and, in parallel, represent both event context and subevent occurrences with learning. Female and male human participants viewed sequential events defined as sequences of objects superimposed on shared color frames while undergoing high-resolution fMRI. Importantly, these events were repeated to induce learning. Event segmentation, as indexed by increased reaction times at event boundaries, was observed in all repetitions. Temporal memory decisions were quicker for items from the same event compared to across different events, indicating that events shaped memory. With learning, hippocampal CA3 multivoxel activation patterns clustered to reflect the event context, with more clustering correlated with behavioral facilitation during event transitions. In contrast, in the dentate gyrus (DG), temporally proximal items that belonged to the same event became associated with more differentiated neural patterns. A computational model explained these results by dynamic inhibition in the DG. Additional similarity measures support the notion that CA3 clustered representations reflect shared voxel populations, while DG's distinct item representations reflect different voxel populations. These findings suggest an interplay between temporal differentiation in the DG and attractor dynamics in CA3. They advance our understanding of how knowledge is structured through integration and separation across time and context. [ABSTRACT FROM AUTHOR]

Early steps in the origin of life were necessarily connected to the unlikely formation of self-reproducing structures from chaotic chemistry. Simulations of chemical kinetics based on the graded autocatalysis replication domain (GARD) model demonstrate the ability of a micellar system to become self-reproducing units away from equilibrium. Even though they may be very rare in the initial state of the system, the property of their endogenous mutually catalytic networks being dynamic attractors greatly enhanced reproduction propensity, revealing their potential for selection and Darwinian evolution processes. In parallel, order and complexity have been shown to be crucial parameters in successful evolution. Here, we probe these parameters in the dynamics of GARD-governed entities in an attempt to identify characteristic mechanisms of their development in non-covalent molecular assemblies. Using a virtual random walk perspective, a value for consecutive order is defined based on statistical thermodynamics. The complexity, on the other hand, is determined by the size of a minimal algorithm fully describing the statistical properties of the random walk. By referring to a previously published diagonal line in an order/complexity diagram that represents the progression of evolution, it is shown that the GARD model has the potential to advance in this direction. These results can serve as a solid foundation for identifying general criteria for future analyses of evolving systems. [ABSTRACT FROM AUTHOR]

Flexibility is a hallmark of memories that depend on the hippocampus. For navigating animals, flexibility is necessitated by environmental changes such as blocked paths and extinguished food sources. To better understand the neural basis of this flexibility, we recorded hippocampal replays in a spatial memory task where barriers as well as goals were moved between sessions to see whether replays could adapt to new spatial and reward contingencies. Strikingly, replays consistently depicted new goal-directed trajectories around each new barrier configuration and largely avoided barrier violations. Barrier-respecting replays were learned rapidly and did not rely on place cell remapping. These data distinguish sharply between place field responses, which were largely stable and remained tied to sensory cues, and replays, which changed flexibly to reflect the learned contingencies in the environment and suggest sequenced activations such as replay to be an important link between the hippocampus and flexible memory.

The mammalian navigation system comprises various kinds of neurons responsible for position perception and spatial path planning, involving the integration of multiple information sources. As a unified brain theory capable of providing explanations for complex cognitive functions like memory and decision-making, the theory of attractor dynamics can elucidate the firing dynamics of neurons and path integration in the navigation system. This review describes recent advances in attract。y dynamics in spatial cognition. First, it provides a brief overview of computational neuroscience and the general theory of attract oy dynamics. Subsequently, focusing on the continuous attractor dynamics, it delves into the dynamical characteristics and functional significance of head direction cells and grid cells. Finally, an extension and prospects of the attractort heory for spatial cognition are presented. [ABSTRACT FROM AUTHOR]

The recent discovery of the head-direction (HD) system in fruit flies has provided unprecedented insights into the neural mechanisms of spatial orientation. Despite the progress, the neural substance of global inhibition, an essential component of the HD circuits, remains controversial. Some studies suggested that the ring neurons provide global inhibition, while others suggested the Δ7 neurons. In the present study, we provide evaluations from the theoretical perspective by performing systematic analyses on the computational models based on the ring-neuron (R models) and Δ7-neurons (Delta models) hypotheses with modifications according to the latest connectomic data. We conducted four tests: robustness, persistency, speed, and dynamical characteristics. We discovered that the two models led to a comparable performance in general, but each excelled in different tests. The R Models were more robust, while the Delta models were better in the persistency test. We also tested a hybrid model that combines both inhibitory mechanisms. While the performances of the R and Delta models in each test are highly parameter-dependent, the Hybrid model performed well in all tests with the same set of parameters. Our results suggest the possibility of combined inhibitory mechanisms in the HD circuits of fruit flies. [ABSTRACT FROM AUTHOR]

Early steps in the origin of life were necessarily connected to the unlikely formation of self-reproducing structures from chaotic chemistry. Simulations of chemical kinetics based on the graded autocatalysis replication domain (GARD) model demonstrate the ability of a micellar system to become self-reproducing units away from equilibrium. Even though they may be very rare in the initial state of the system, the property of their endogenous mutually catalytic networks being dynamic attractors greatly enhanced reproduction propensity, revealing their potential for selection and Darwinian evolution processes. In parallel, order and complexity have been shown to be crucial parameters in successful evolution. Here, we probe these parameters in the dynamics of GARD-governed entities in an attempt to identify characteristic mechanisms of their development in non-covalent molecular assemblies. Using a virtual random walk perspective, a value for consecutive order is defined based on statistical thermodynamics. The complexity, on the other hand, is determined by the size of a minimal algorithm fully describing the statistical properties of the random walk. By referring to a previously published diagonal line in an order/complexity diagram that represents the progression of evolution, it is shown that the GARD model has the potential to advance in this direction. These results can serve as a solid foundation for identifying general criteria for future analyses of evolving systems.

This paper outlines a framework for strength training as a dynamical model of perceptual-motor learning. We show, with emphasis on fixed-point attractor dynamics, that strength training can be mapped to the general dynamical principles of motor learning that arise from the constraints on action, including the distribution of practice/training. The time scales of the respective dynamics of performance change (increment and decrement) in discrete strength training and motor learning tasks reveal superposition of exponential functions in fixed-point dynamics, but distinctive attractor and parameter dynamics in oscillatory limit cycle and more continuous tasks, together with unique timescales to process influences (including practice, learning, strength, fitness, fatigue, warm-up decrement). Increments and decrements of strength can be viewed within a dynamical model of change in motor performance that reflects the integration of practice and training processes at multiple levels of learning and skill development. [ABSTRACT FROM AUTHOR]

The preBötzinger Complex (preBötC) encodes inspiratory time as rhythmic bursts of activity underlying each breath. Spike synchronization throughout a sparsely connected preBötC microcircuit initiates bursts that ultimately drive the inspiratory motor patterns. Using minimal microcircuit models to explore burst initiation dynamics, we examined the variability in probability and latency to burst following exogenous stimulation of a small subset of neurons, mimicking experiments. Among various physiologically plausible graphs of 1000 excitatory neurons constructed using experimentally determined synaptic and connectivity parameters, directed Erdoös-Rényi graphs with a broad (lognormal) distribution of synaptic weights best captured the experimentally observed dynamics. preBötC synchronization leading to bursts was regulated by the efferent connectivity of spiking neurons that are optimally tuned to amplify modest preinspiratory activity through input convergence. Using graph-theoretic and machine learning-based analyses, we found that input convergence of efferent connectivity at the next nearest neighbor order was a strong predictor of incipient synchronization. Our analyses revealed a crucial role of synaptic heterogeneity in imparting exceptionally robust yet flexible preBötC attractor dynamics. Given the pervasiveness of lognormally distributed synaptic strengths throughout the nervous system, we postulate that these mechanisms represent a ubiquitous template for temporal processing and decision-making computational motifs. [ABSTRACT FROM AUTHOR]

Cingulotomy is therapeutic in OCD, but what are the possible mechanisms? Computer models that formalize cortical OCD abnormalities and anterior cingulate cortex (ACC) function can help answer this. At the neural dynamics level, cortical dynamics in OCD have been modeled using attractor networks, where activity patterns resistant to change denote the inability to switch to new patterns, which can reflect inflexible thinking patterns or behaviors. From that perspective, cingulotomy might reduce the influence of difficult-to-escape ACC attractor dynamics on other cortical areas. At the functional level, computer formulations based on model-free reinforcement learning (RL) have been used to describe the multitude of phenomena ACC is involved in, such as tracking the timing of expected outcomes and estimating the cost of exerting cognitive control and effort. Different elements of model-free RL models of ACC could be affected by the inflexible cortical dynamics, making it challenging to update their values. An agent can also use a world model, a representation of how the states of the world change, to plan its actions, through model-based RL. OCD has been hypothesized to be driven by reduced certainty of how the brain's world model describes changes. Cingulotomy might improve such uncertainties about the world and one's actions, making it possible to trust the outcomes of these actions more and thus reduce the urge to collect more sensory information in the form of compulsions. Connecting the neural dynamics models with the functional formulations can provide new ways of understanding the role of ACC in OCD, with potential therapeutic insights. [ABSTRACT FROM AUTHOR]

Here we present an analysis of eigenvalues and eigenvectors of a correlation matrix that contains intra- and inter-frequency band correlations for narrow band filtered EEG signals of healthy subjects during sleep. We find that inter-band correlations of the same electrode are particularly suitable for distinguishing different sleep stages. However, largest eigenvalues and their corresponding eigenvectors show an extremely stable behavior during night sleep, while sleep stages manifest via specific deviations from an overall stable structure expressed by relative eigenvalues and deviations from average eigenvectors. Concluding that valuable information is encoded along the whole spectrum of eigenvalues and eigenvectors of the correlation matrix, we interpret our results in term of dynamical system theory. • Eigenvalues and eigenvectors correlation matrices from EEG during sleep reveal distinct patterns. • Inter-band correlations are effective in distinguishing between different sleep stages. • Largest eigenvalues and their corresponding eigenvectors exhibit stable behavior throughout night sleep. • Sleep stages are characterized by deviations from this stable structure. [ABSTRACT FROM AUTHOR]

Cingulotomy is therapeutic in OCD, but what are the possible mechanisms? Computer models that formalize cortical OCD abnormalities and anterior cingulate cortex (ACC) function can help answer this. At the neural dynamics level, cortical dynamics in OCD have been modeled using attractor networks, where activity patterns resistant to change denote the inability to switch to new patterns, which can reflect inflexible thinking patterns or behaviors. From that perspective, cingulotomy might reduce the influence of difficult-to-escape ACC attractor dynamics on other cortical areas. At the functional level, computer formulations based on model-free reinforcement learning (RL) have been used to describe the multitude of phenomena ACC is involved in, such as tracking the timing of expected outcomes and estimating the cost of exerting cognitive control and effort. Different elements of model-free RL models of ACC could be affected by the inflexible cortical dynamics, making it challenging to update their values. An agent can also use a world model, a representation of how the states of the world change, to plan its actions, through model-based RL. OCD has been hypothesized to be driven by reduced certainty of how the brain’s world model describes changes. Cingulotomy might improve such uncertainties about the world and one’s actions, making it possible to trust the outcomes of these actions more and thus reduce the urge to collect more sensory information in the form of compulsions. Connecting the neural dynamics models with the functional formulations can provide new ways of understanding the role of ACC in OCD, with potential therapeutic insights.

Temporal accumulation of evidence is crucial for making accurate judgments based on noisy or ambiguous sensory input. The integration process leading to categorical decisions is thought to rely on competition between neural populations, each encoding a discrete categorical choice. How recurrent neural circuits integrate evidence for continuous perceptual judgments is unknown. Here, we show that a continuous bump attractor network can integrate a circular feature, such as stimulus direction, nearly optimally. As required by optimal integration, the population activity of the network unfolds on a twodimensional manifold, in which the position of the network’s activity bump tracks the stimulus average, and, simultaneously, the bump amplitude tracks stimulus uncertainty. Moreover, the temporal weighting of sensory evidence by the network depends on the relative strength of the stimulus compared to the internally generated bump dynamics, yielding either early (primacy), uniform, or late (recency) weighting. The model can flexibly switch between these regimes by changing a single control parameter, the global excitatory drive. We show that this mechanism can quantitatively explain individual temporal weighting profiles of human observers, and we validate the model prediction that temporal weighting impacts reaction times. Our findings point to continuous attractor dynamics as a plausible neural mechanism underlying stimulus integration in perceptual estimation tasks. [ABSTRACT FROM AUTHOR]

Distinct computations are performed at multiple brain regions during the encoding of spatial environments. Neural representations in the hippocampal, entorhinal, and head direction (HD) networks during spatial navigation have been clearly docu- mented, while the representational properties of the subicular complex (SC) are relatively underexplored, although it has extensive anatomic connections with various brain regions involved in spatial information processing. We simultaneously recorded single units from different subregions of the SC in male rats while they ran clockwise on a centrally placed textured circular track (four different textures, each covering a quadrant), surrounded by six distal cues. The neural activity was monitored in standard sessions by maintaining the same configuration between the cues, while in cue manipulation sessions, the distal and local cues were either rotated in opposite directions to create a mismatch between them or the distal cues were removed. We report a highly coherent neural representation of the environment and a robust coupling between the HD cells and the spatial cells in the SC, strikingly different from previous reports of coupling between cells from co-recorded sites. Neural representations were (1) originally governed by the distal cues under local--distal cue-conflict conditions, (2) controlled by the local cues in the absence of distal cues, and (3) governed by the cues that are perceived to be stable. We propose that such attractor-like dynamics in the SC might play a critical role in the orientation of spatial representations, thus providing a "reference map" of the environment for further processing by other networks. [ABSTRACT FROM AUTHOR]

Errors in stimulus estimation reveal how stimulus representation changes during cognitive processes. Repulsive bias and minimum variance observed near cardinal axes are well-known error patterns typically associated with visual orientation perception. Recent experiments suggest that these errors continuously evolve during working memory, posing a challenge that neither static sensory models nor traditional memory models can address. Here, we demonstrate that these evolving errors, maintaining characteristic shapes, require network interaction between two distinct modules. Each module fulfills efficient sensory encoding and memory maintenance, which cannot be achieved simultaneously in a single-module network. The sensory module exhibits heterogeneous tuning with strong inhibitory modulation reflecting natural orientation statistics. While the memory module, operating alone, supports homogeneous representation via continuous attractor dynamics, the fully connected network forms discrete attractors with moderate drift speed and nonuniform diffusion processes. Together, our work underscores the significance of sensory-memory interaction in continuously shaping stimulus representation during working memory., Competing Interests: JY, HZ, SL No competing interests declared, (© 2024, Yang et al.)

Autoassociative neural networks provide a simple model of how memories can be stored through Hebbian synaptic plasticity as retrievable patterns of neural activity. Although progress has been made along the last decades in understanding the biological implementation of autoassociative networks, their modest theoretical storage capacity has remained a major constraint. While most previous approaches utilize randomly connected networks, here we explore the possibility of optimizing network performance by selective connectivity between neurons, that could be implemented in the brain through creation and pruning of synaptic connections. We show through numerical simulations that a reconfiguration of the connectivity matrix can improve the storage capacity of autoassociative networks up to one order of magnitude compared to randomly connected networks, either by reducing the noise or by making it reinforce the signal. Our results indicate that the signal-reinforcement scenario is not only the best performing but also the most adequate for brain-like highly diluted connectivity. In this scenario, the optimized network tends to select synapses characterized by a high consensus across stored patterns. We also introduced an online algorithm in which the network modifies its connectivity while learning new patterns. We observed that, similarly to what happens in the human brain, creation of connections dominated in an initial stage, followed by a stage characterized by pruning, leading to an equilibrium state that was independent of the initial connectivity of the network. Our results suggest that selective connectivity could be a key component to make attractor networks in the brain viable in terms of storage capacity. [ABSTRACT FROM AUTHOR]

The outcomes of intertemporal choices indicate that peoplediscount rewards by their delay. These outcomes are welldescribed by discounting functions. However, to fullyunderstand the decision process one needs models describinghow the process of decision-making unfolds dynamically overtime. Here, we validate a recently published attractor modelthat extends discounting functions through a description ofthe dynamics leading to a final choice outcome within andacross trials. We focus on the decision dynamics across trials.We derive qualitative predictions for the inter-trial dynamicsof sequences of decisions that are unique to this type ofmodel. We test these predictions in a delay discounting gamewhere we sequentially manipulated subjective values ofoptions across all attribute dimensions. Results confirm themodel’s predictions. We discuss future challenges onintegrating attractor models towards a general attractor modelof delay discounting to enhance our understanding of theprocesses underlying delay discounting decisions.

There is strong evidence that anesthetics have stereotypical effects on brain state, so that a given anesthetic appears to have a signature in the electroencephalogram (EEG), which may vary with dose. This can be usefully interpreted as the anesthetic determining an attractor in the phase space of the brain. How brain activity shifts between these attractors in time remains understudied, as most studies implicitly assume a one-to-one relationship between drug dose and attractor features by assuming stationarity over the analysis interval and analyzing data segments of several minutes in length. Yet data in rats anesthetized with isoflurane suggests that, at anesthetic levels consistent with surgical anesthesia, brain activity alternates between multiple attractors, often spending on the order of 10 min in one activity pattern before shifting to another. Moreover, the probability of these jumps between attractors changes with anesthetic concentration. This suggests the hypothesis that brain state is metastable during anesthesia: though it appears at equilibrium on short timescales (on the order of seconds to a few minutes), longer intervals show shifting behavior. Compelling evidence for metastability in rats anesthetized with isoflurane is reviewed, but so far only suggestive hints of metastability in brain states exist with other anesthetics or in other species. Explicit testing of metastability during anesthesia will require experiments with longer acquisition intervals and carefully designed analytic approaches; some of the implications of these constraints are reviewed for typical spectral analysis approaches. If metastability exists during anesthesia, it implies degeneracy in the relationship between brain state and effect site concentration, as there is not a one-to-one mapping between the two. This degeneracy could explain some of the reported difficulty in using brain activity monitors to titrate drug dose to prevent awareness during anesthesia and should force a rethinking of the notion of depth of anesthesia as a single dimension. Finally, explicit incorporation of knowledge of the dynamics of the brain during anesthesia could offer better depth of anesthesia monitoring.

Sequences of neural activity arise in many brain areas, including cortex, hippocampus, and central pattern generator circuits that underlie rhythmic behaviors like locomotion. While network architectures supporting sequence generation vary considerably, a common feature is an abundance of inhibition. In this work, we focus on architectures that support sequential activity in recurrently connected networks with inhibition-dominated dynamics. Specifically, we study emergent sequences in a special family of threshold-linear networks, called combinatorial threshold-linear networks (CTLNs), whose connectivity matrices are defined from directed graphs. Such networks naturally give rise to an abundance of sequences whose dynamics are tightly connected to the underlying graph. We find that architectures based on generalizations of cycle graphs produce limit cycle attractors that can be activated to generate transient or persistent (repeating) sequences. Each architecture type gives rise to an infinite family of graphs that can be built from arbitrary component subgraphs. Moreover, we prove a number of graph rules for the corresponding CTLNs in each family. The graph rules allow us to strongly constrain, and in some cases fully determine, the fixed points of the network in terms of the fixed points of the component subnetworks. Finally, we also show how the structure of certain architectures gives insight into the sequential dynamics of the corresponding attractor. [ABSTRACT FROM AUTHOR]

A continuous attractor neural network (CANN) configured with various synaptic dynamics can implement many brain cognitive functions. Corresponding neuromorphic implementation can be beneficial to carry out these cognitive computations in real-time. Herein, we present a neuromorphic implementation of the CANN of spiking neurons on the field-programmable gate array (FPGA), in which the synaptic strength of local excitation and global inhibition can be configured to achieve winner-take-all (WTA) competition. In contrast to previously reported neuromorphic CANN, in addition to fast-linear synapses, recurrent connections can also be implemented with slow-nonlinear synapses. This endows the proposed CANN on the FPGA with the capability of performing neural computations that require slow and nonlinear temporal dynamics, such as decision making and working memory. Circuit simulations on the hardware reproduced gradually ramping neural activities observed in decision-making experiments and multiple sustained activities observed during the delay in working memory experiments, especially the fade-out and merging of activity bumps in the latter task. Theoretical analysis illuminated the underlying mechanisms of decision making and working memory on the FPGA. The unstable two-bump steady state of the CANN is responsible for the two-alternative decision making, similar to the saddle-node structure in the reduced two-variable decision model. Meanwhile, the stable uniformly distributed N-bump steady states account for the multiple-item working memory, similar to the multistability in previous mean-field models. Furthermore, the asymmetrical N-bump unstable steady states exhibit attractor dynamics underlying the fade-out and merging of activity bumps. The consistency between simulations on the hardware and the theoretical analysis demonstrated that the neuromorphic CANN is endowed with attractor dynamics relating to slow and nonlinear neural computations and has the potential to implement sophisticated cognitive functions by configuring various synaptic dynamics. This implementation shows promise for integrating the cognitive platform due to its characteristics of various dynamics and flexible configurations.

In ambiguous or conflicting sensory situations, perception is often ‘multistable’ in that it perpetually changes at irregular intervals, shifting abruptly between distinct alternatives. The interval statistics of these alternations exhibits quasi-universal characteristics, suggesting a general mechanism. Using binocular rivalry, we show that many aspects of this perceptual dynamics are reproduced by a hierarchical model operating out of equilibrium. The constitutive elements of this model idealize the metastability of cortical networks. Independent elements accumulate visual evidence at one level, while groups of coupled elements compete for dominance at another level. As soon as one group dominates perception, feedback inhibition suppresses supporting evidence. Previously unreported features in the serial dependencies of perceptual alternations compellingly corroborate this mechanism. Moreover, the proposed out-of-equilibrium dynamics satisfies normative constraints of continuous decision-making. Thus, multistable perception may reflect decision-making in a volatile world: integrating evidence over space and time, choosing categorically between hypotheses, while concurrently evaluating alternatives.

Traditionally, resilience has been viewed as a general positive adaptation to stressors. However, the hallmark of resilience – returning to the previous state following a perturbation – may also have severe downsides, which are often overlooked. Specifically, it may be unrealistic to return to the previous state or resilience may cause a person to become stuck in an undesirable state. In this article, we first call for a more nuanced theoretical conceptualization of resilience. To do so, we draw on insights from dynamical systems theory help to clearly define the role of a stressor and the idealized pathway to adapt to it. Then, we exemplify the potential downsides of resilience in the context of trauma and social adversity, learning, and goal-disengagement. In conclusion, researchers and practitioners should become more cautious with the term resilience and provide nuanced accounts for what they mean to avoid potentially harmful consequences. • Resilience – defined as the process of bouncing back – is not always desirable. • Resilience can cause a person to become stuck in an undesirable state. • Practitioners need to be cautious with overly emphasizing resilience. • Dynamical systems theory can help to build a clearer understanding of resilience. [ABSTRACT FROM AUTHOR]

The precise coordination of neuronal activity is critical for optimal brain function. When such coordination fails, this can lead to dire consequences. In this review, I will present evidence that in epilepsy, failed coordination leads not only to seizures but also to alterations of the rhythmical patterns observed in the electroencephalogram and cognitive deficits. Restoring the dynamic coordination of epileptic networks could therefore both improve seizures and cognitive outcomes. [ABSTRACT FROM AUTHOR]

In real life, decisions are often naturally embedded in decision sequences. In contrast, in the laboratory, decisions are oftentimes analysed in isolation. Here, we investigated the influence of decision sequences in value-based decision making and whether the stability of such effects can be modulated. In our decision task, participants needed to collect rewards in a virtual two-dimensional world. We presented a series of two reward options that were either quick to collect but were smaller in value or took longer to collect but were larger in value. The subjective value of each option was driven by the options' value and how quickly they could be reached. We manipulated the subjective values of the options so that one option became gradually less valuable over the course of a sequence, which allowed us to measure choice perseveration (i.e., how long participants stick to this option). In two experiments, we further manipulated the time interval between two trials (inter-trial interval), and the time delay between the onsets of both reward options (stimulus onset asynchrony). We predicted how these manipulations would affect choice perseveration using a computational attractor model. Our results indicate that both the inter-trial interval and the stimulus onset asynchrony modulate choice perseveration as predicted by the model. We discuss how our findings extend to research on cognitive stability and flexibility. [ABSTRACT FROM AUTHOR]

Spatial navigation depends on the combination of multiple sensory cues from idiothetic and allothetic sources. The computational mechanisms of mammalian brains in integrating different sensory modalities under uncertainty for navigation is enlightening for robot navigation. We propose a Bayesian attractor network model to integrate visual and vestibular inputs inspired by the spatial memory systems of mammalian brains. In the model, the pose of the robot is encoded separately by two sub-networks, namely head direction network for angle representation and grid cell network for position representation, using similar neural codes of head direction cells and grid cells observed in mammalian brains. The neural codes in each of the sub-networks are updated in a Bayesian manner by a population of integrator cells for vestibular cue integration, as well as a population of calibration cells for visual cue calibration. The conflict between vestibular cue and visual cue is resolved by the competitive dynamics between the two populations. The model, implemented on a monocular visual simultaneous localization and mapping (SLAM) system, termed NeuroBayesSLAM, successfully builds semi-metric topological maps and self-localizes in outdoor and indoor environments of difference characteristics, achieving comparable performance as previous neurobiologically inspired navigation systems but with much less computation complexity. The proposed multisensory integration method constitutes a concise yet robust and biologically plausible method for robot navigation in large environments. The model provides a viable Bayesian mechanism for multisensory integration that may pertain to other neural subsystems beyond spatial cognition. • A Bayesian attractor network model emulating neural encoding mechanisms of head direction cells and grid cells of the mammalian spatial memory circuits. • Bayesian integration of vestibular cues by integrator cells and visual cues by calibration cells separately. • Cue conflict resolution by competitive dynamics between integrator cells and calibration cells. • Robust performance in indoor and large outdoor environments. [ABSTRACT FROM AUTHOR]

Because the input and the output values of the deep denoising autoencoders (DDAs) have the same representation space, the output values of a DDA can be used as its input values, which leads to a repeatable cycle for DDA. Using this fact, in this paper we proposed cyclic deep denoising autoencoders (CDDAs). Moreover, the proposed CDDA was exploited for noise robust bottleneck feature extraction. Considering the CDDA as an attractor network, we analyzed its attractor dynamics and proposed an uncertainty estimation method for the bottleneck features. The capabilities of the proposed feature extraction method are: eliminating the need for labeled training data, improving the robustness of bottleneck features of DDAs without excessive training, providing a new uncertainty estimation trend, and being applicable in other speech processing tasks like speaker recognition. Our proposed feature extraction method was evaluated on the Aurora-2 robust speech recognition task and compared it to conventional Mel frequency cepstral coefficients (MFCC) and autoencoder bottleneck features. The proposed method resulted in up to 8% improvement in word recognition accuracy relative to MFCC features. [ABSTRACT FROM AUTHOR]

This paper shows how non-linear attractor dynamics can be used to control teams of two autonomous mobile robots that coordinate their motion in order to transport large payloads in unknown environments, which might change over time and may include narrow passages, corners and sharp U-turns. Each robot generates its collision-free motion online as the sensed information changes. The control architecture for each robot is formalized as a non-linear dynamical system, where by design attractor states, i.e. asymptotically stable states, dominate and evolve over time. Implementation details are provided, and it is further shown that odometry or calibration errors are of no significance. Results demonstrate flexible and stable behavior in different circumstances: when the payload is of different sizes; when the layout of the environment changes from one run to another; when the environment is dynamic—e.g. following moving targets and avoiding moving obstacles; and when abrupt disturbances challenge team behavior during the execution of the joint transportation task. [ABSTRACT FROM AUTHOR]

How do people make decisions between an immediate but small reward and a delayed but large one? The outcome of such decisions indicates that people discount rewards by their delay and hence these outcomes are well described by discounting functions. However, to understand irregular decisions and dysfunctional behavior one needs models which describe how the process of making the decision unfolds dynamically over time: how do we reach a decision and how do sequential decisions influence one another? Here, we present an attractor model that integrates into and extends discounting functions through a description of the dynamics leading to a final choice outcome within a trial and across trials. To validate this model, we derive qualitative predictions for the intra-trial dynamics of single decisions and for the inter-trial dynamics of sequences of decisions that are unique to this type of model. We test these predictions in four experiments based on a dynamic delay discounting computer game where we study the intra-trial dynamics of single decisions via mouse tracking and the inter-trial dynamics of sequences of decisions via sequentially manipulated options. We discuss how integrating decision process dynamics within and across trials can increase our understanding of the processes underlying delay discounting decisions and, hence, complement our knowledge about decision outcomes.

Fast object tracking on embedded devices is of great importance for applications such as autonomous driving, unmanned aerial vehicle, and intelligent monitoring. Whereas, most of previous general solutions failed to reach this goal due to the facts that (i) high computational complexity and heterogeneous operation steps in the tracking models and (ii) parallelism-limited and bloated hardware platforms (e.g., CPU/GPU). Although previously proposed devices leverage neural dynamics and near-data processing for efficient tracking, their flexibility is limited due to the tight integration with vision sensor and the effectiveness on various video datasets is yet to be fully demonstrated. On the other side, recently the many-core architecture with massive parallelism and optimized memory locality is being widely applied to improve the performance for flexibly executing neural networks. This motivates us to adapt and map an object tracking model based on attractor neural networks with continuous and smooth attractor dynamics onto neural network chips for fast tracking. In order to make the model hardware friendly, we add local-connection restriction. We analyze the tracking accuracy and observe that the model achieves comparable results on typical video datasets. Then, we design a many-core neural network architecture with several computation and transformation operations to support the model. Moreover, by discretizing the continuous dynamics to the corresponding discrete counterpart, designing a slicing scheme for efficient topology mapping, and introducing a constant-restricted scaling chain rule for data quantization, we build a complete mapping framework to implement the tracking model on the many-core architecture. We fabricate a many-core neural network chip to evaluate the real execution performance. Results show that a single chip is able to accommodate the whole tracking model, and a fast tracking speed of nearly 800 FPS (frames per second) can be achieved. This work enables high-speed object tracking on embedded devices which normally have limited resources and energy.

Maintaining a sense of direction is fundamental to navigation, and is achieved in the brain by a network of head direction (HD) cells, which update their signal using stable environmental landmarks. How landmarks are detected and their stability determined is still unknown. Recently we reported a new class of cells (Jacob et al., 2017), the bidirectional cells, in a brain region called retrosplenial cortex (RSC) which relays environmental sensory information to the HD system. A subset of these cells, between-compartment (BC) cells, are directionally tuned (like ordinary HD cells) but follow environmental cues in preference to the global HD signal, resulting in opposing (i.e., bidirectional) tuning curves in opposed environments. Another subset, within-compartment (WC) cells, unexpectedly expressed bidirectional tuning curves in each one of the opposed compartments. Both BC and WC cells lost directional tuning in an open field, unlike HD cells. Two questions arise from this discovery: (i) how do these cells acquire their unusual response properties, and (ii) what are they for? We propose that bidirectional cells reflect a two-way interaction between local direction, as indicated by the visual environment, and global direction as signaled by the HD system. We suggest that BC cells receive strong inputs from visual cues, while WC cells additionally receive modifiable inputs from HD cells which, due to Hebbian coactivation of visual inputs plus two opposing sets of HD inputs, acquire the ability to fire in both directions. A neural network model instantiating this hypothesis is presented, which indeed forms both BC and WC bidirectional cells with properties similar to those seen experimentally. We then demonstrate how tuning specificity degrades when WC/BC cells are exposed to multiple directionalities, replicating the observed loss of WC and BC directional tuning in the open field. We suggest that the function of these neurons is to assess the stability of environmental landmarks, thereby determining their utility as reference points by which to set the HD sense of direction. This role could extend to the ability of the HD system to prefer distal over proximal landmarks, and to correct for parallax errors.

A goal of systems neuroscience is to discover the circuit mechanisms underlying brain function. Despite experimental advances that enable circuit-wide neural recording, the problem remains open in part because solving the ‘inverse problem’ of inferring circuity and mechanism by merely observing activity is hard. In the grid cell system, we show through modeling that a technique based on global circuit perturbation and examination of a novel theoretical object called the distribution of relative phase shifts (DRPS) could reveal the mechanisms of a cortical circuit at unprecedented detail using extremely sparse neural recordings. We establish feasibility, showing that the method can discriminate between recurrent versus feedforward mechanisms and amongst various recurrent mechanisms using recordings from a handful of cells. The proposed strategy demonstrates that sparse recording coupled with simple perturbation can reveal more about circuit mechanism than can full knowledge of network activity or the synaptic connectivity matrix.