Dynamics control of semantic processes in a hierarchical associative memory

 A neural mechanism for control of dynamics and function of associative processes in a hierarchical memory system is demonstrated. For the representation and processing of abstract knowledge, the semantic declarative memory system of the human brain is considered. The dynamics control mechanism is based on the influence of neuronal adaptation on the complexity of neural network dynamics. Different dynamical modes correspond to different levels of the ultrametric structure of the hierarchical memory being invoked during an associative process. The mechanism is deterministic but may also underlie free associative thought processes. The formulation of an abstract neural network model of hierarchical associative memory utilizes a recent approach to incorporate neuronal adaptation. It includes a generalized neuronal activation function recently derived by a Hodgkin-Huxley-type model. It is shown that the extent to which a hierarchically organized memory structure is searched is controlled by the neuronal adaptability, i.e. the strength of coupling between neuronal activity and excitability. In the brain, the concentration of various neuromodulators in turn can regulate the adaptability. An autonomously controlled sequence of bifurcations, from an initial exploratory to a final retrieval phase, of an associative process is shown to result from an activity-dependent release of neuromodulators. The dynamics control mechanism may be important in the context of various disorders of the brain and may also extend the range of applications of artificial neural networks. Received: 19 April 1995/Accepted in revised form: 8 August 1995     

Towards an artificial brain

Three components of a brain model operating on neuromolecular computing principles are described. The first component comprises neurons whose input-output behavior is controlled by significant internal dynamics. Models of discrete enzymatic neurons, reaction-diffusion neurons operating on the basis of the cyclic nucleotide cascade, and neurons controlled by cytoskeletal dynamics are described. The second component of the model is an evolutionary learning algorithm which is used to mold the behavior of enzyme-driven neurons or small networks of these neurons for specific function, usually pattern recognition or target seeking tasks. The evolutionary learning algorithm may be interpreted either as representing the mechanism of variation and natural selection acting on a phylogenetic time scale, or as a conceivable ontogenetic adaptation mechanism. The third component of the model is a memory manipulation scheme, called the reference neuron scheme. In principle it is capable of orchestrating a repertoire of enzyme-driven neurons for coherent function. The existing implementations, however, utilize simple neurons without internal dynamics. Spatial navigation and simple game playing (using tic-tac-toe) provide the task environments that have been used to study the properties of the reference neuron model. A memory-based evolutionary learning algorithm has been developed that can assign credit to the individual neurons in a network. It has been run on standard benchmark tasks, and appears to be quite effective both for conventional neural nets and for networks of discrete enzymatic neurons. The models have the character of artificial worlds in that they map the hierarchy of processes in the brain (at the molecular, neuronal, and network levels), provide a task environment, and use this relatively self-contained setup to develop and evaluate learning and adaptation algorithms.     

Neuromodulatory Control of Neocortical Microcircuits with Activity-Dependent Short-Term Synaptic Depression

A biophysical model of a neocortical microcircuit system is formulated and employed in studies of neuromodulatory control of dynamics and function. The model is based on recent observations of reciprocal connections between pyramidal cells and inhibitory interneurons and incorporates a new type of activity-dependent short-term depression of synaptic couplings recently observed. The model neurons are of a low-dimensional type also accounting for neuronal adaptation, i.e. the coupling between neuronal activity and excitability, which can be regulated by various neuromodulators in the brain. The results obtained demonstrate a capacity for neuromodulatory control of dynamical mode linked to functional mode. The functional aspects considered refer to the observed resolution of multiple objects in working memory as well as the binding of different features for the perception of an object. The effects of neuromodulators displayed by the model are in accordance with many observations on neuromodulatory influence on cognitive functions and brain disorders.     

Control of computational dynamics of coupled integrate-and-fire neurons

 Generation and control of different dynamical modes of computational processes in a net of interconnected integrate-and-fire neurons are demonstrated. A net architecture resembling a generic cortical structure is formed from pairs of excitatory and inhibitory units with excitatory connections between and inhibitory connections within pairs. Integrate-and-fire model neurons derived from detailed conductance-based models of neocortical pyramidal cells and fast-spiking interneurons are employed for the excitatory and inhibitory units, respectively. Firing-rate adaptation is incorporated into the excitatory units based on the regulation of the slow afterhyperpolarization phase of action potentials by intracellular calcium ions. Saturation of synaptic conductances is implemented for the interconnections between units. It is shown that neuronal adaptation of the excitatory units can generate richer net dynamics than relaxation to fixed-point attractors in a pattern space. At strong adaptivity, i.e. when the neuronal excitability is strongly influenced by the preceding activity, complex dynamics of either aperiodic or limit-cycle character are generated in both the pattern space and the phase space of all dynamical variables. This regime corresponds to an exploratory mode of the system, in which the pattern space can be searched. At weak adaptivity, the dynamics are governed by fixed-point attractors in the pattern space, and this corresponds to a mode for retrieval of a particular pattern. In the brain, neuronal adaptivity can be regulated by various neuromodulators. The results are in accordance with those recently obtained by means of more abstract models formulated in terms of mean firing rates. The increased realism makes the present model reveal more detailed mechanisms and strengthens the relevance of the conclusions to biological systems. The simplicity and realism of the coupled integrate-and-fire neurons make the present model useful for studies of systems in which the temporal aspects of neural coding are important. Received: 8 December 1995 / Accepted in revised form: 23 January 1997     

Self-organization of associative memory and pattern classification: recurrent signal processing on topological feature maps

We extend the neural concepts of topological feature maps towards self-organization of auto-associative memory and hierarchical pattern classification. As is well-known, topological maps for statistical data sets store information on the associated probability densities. To extract that information we introduce a recurrent dynamics of signal processing. We show that the dynamics converts a topological map into an auto-associative memory for real-valued feature vectors which is capable to perform a cluster analysis. The neural network scheme thus developed represents a generalization of non-linear matrix-type associative memories. The results naturally lead to the concept of a feature atlas and an associated scheme of self-organized, hierarchical pattern classification.     

Multiscale representations of community structures in attractor neural networks

Our cognition relies on the ability of the brain to segment hierarchically structured events on multiple scales. Recent evidence suggests that the brain performs this event segmentation based on the structure of state-transition graphs behind sequential experiences. However, the underlying circuit mechanisms are poorly understood. In this paper we propose an extended attractor network model for graph-based hierarchical computation which we call the Laplacian associative memory. This model generates multiscale representations for communities (clusters) of associative links between memory items, and the scale is regulated by the heterogenous modulation of inhibitory circuits. We analytically and numerically show that these representations correspond to graph Laplacian eigenvectors, a popular method for graph segmentation and dimensionality reduction. Finally, we demonstrate that our model exhibits chunked sequential activity patterns resembling hippocampal theta sequences. Our model connects graph theory and attractor dynamics to provide a biologically plausible mechanism for abstraction in the brain.     

Fluctuations and synchronizations of neural activities during sleep: Neural basis of possible sleep functions?

Physiological and psychological evidence have been accumulated concerning the function of sleep in development and learning/memory. Many conceptual ideas have been proposed to elucidate the mechanisms underlying them. Sleep consists of a wide variety of physiological processes. It has not yet been clarified which processes are involved in development and learning/memory processes. We have found that single neuronal activity exhibits a slowly fluctuating rate of discharge during rapid eye movement (REM) sleep and a random low discharge rate during non-rapid eye movement (NREM) sleep. It is suggested that a structural change of the neural network attractor underlies this neuronal dynamics-alternation by mathematical modeling. Functional interpretation of the neuronal dynamics-alternation was provided in combination with the phase locking of ponto-geniculo-occipital (PGO)/pontine (P) wave to the hippocampal theta wave, each of which is known to be involved in learning/memory processes. More directly, by the long-term sensory deprivation, the dynamics of neural activity during sleep was found to progressively change in a non-monotonic way. This finding reveals a possible interaction between sleep and reorganization of neural network in the matured brain. Here, in addition to the related findings, we described our idea about how sleep contributes to the learning/memory processes and reorganization of neural network of the matured brain through characteristic neural activities during sleep.

     

Abstract and associatively based representations in human sequence learning

We give an analysis of performance in an artificial neural network for which the claim had been made that it could learn abstract representations. Our argument is that this network is associative in nature, and cannot develop abstract representations. The network thus converges to a solution that is solely based on the statistical regularities of the training set. Inspired by human experiments that have shown that humans can engage in both associative (statistical) and abstract learning, we present a new, hybrid computational model that combines associative and more abstract, cognitive processes. To cross-validate the model we attempted to predict human behaviour in further experiments. One of these experiments reveals some evidence for the use of abstract representations, whereas the others provide evidence for associatively based performance. The predictions of the hybrid model stand in line with our empirical data.     

Izhikevich神经元网络的同步与联想记忆

联想记忆是人脑的一项重要功能。以Izhikevich神经元模型为节点,构建神经网络,神经元之间采用全连结的方式;以神经元群体的时空编码（spatio-temporal coding）理论研究所构建神经网络的联想记忆功能。在加入高斯白噪声的情况下,调节网络中神经元之间的连接强度的大小,当连接强度和噪声强度达到一个阈值时网络中部分神经元同步放电,实现了存储模式的联想记忆与恢复。仿真结果表明,神经元之间的连接强度在联想记忆的过程中发挥了重要的作用,噪声可以促使神经元间的同步放电,有助于神经网络实现存储模式的联想记忆与恢复。     

Simulation study on dynamics transition in neuronal activity during sleep cycle by using asynchronous and symmetry neural network model

We have found that single neuronal activities in different regions in the brain commonly exhibit the distinct dynamics transition during sleep-waking cycle in cats. Especially, power spectral densities of single neuronal activities change their profiles from the white to the 1/f along with sleep cycle from slow wave sleep (SWS) to paradoxical sleep (PS). Each region has different neural network structure and physiological function. This suggests a globally working mechanism may be underlying the dynamics transition we concern. Pharmacological studies have shown that a change in a wide-spread serotonergic input to these regions possibly causes the neuronal dynamics transition during sleep cycle. In this paper, based on these experimental results, an asynchronous and symmetry neural network model including inhibitory input, which represents the role of the serotonergic system, is utilized to examine the reality of our idea that the inhibitory input level varying during sleep cycle induce that transition. Simulation results show that the globally applied inhibitory input can control the dynamics of single neuronal state evolution in the artificial neural network: 1/f-like power spectral density profiles result under weak inhibition, which possibly corresponds to PS, and white profiles under strong inhibition, which possibly corresponds to SWS. An asynchronous neural network is known to change its state according to its energy function. The geometrical structure of network energy function is thought to vary along with the change in inhibitory level, which is expected to cause the dynamics transition of neuronal state evolution in the network model. These simulation results support the possibility that the serotonergic system is essential for the dynamics transition of single neuronal activities during sleep cycle.     

Spatial-area selective retrieval of multiple object–place associations in a hierarchical cognitive map formed by theta phase coding

The human cognitive map is known to be hierarchically organized consisting of a set of perceptually clustered landmarks. Patient studies have demonstrated that these cognitive maps are maintained by the hippocampus, while the neural dynamics are still poorly understood. The authors have shown that the neural dynamic “theta phase precession” observed in the rodent hippocampus may be capable of forming hierarchical cognitive maps in humans. In the model, a visual input sequence consisting of object and scene features in the central and peripheral visual fields, respectively, results in the formation of a hierarchical cognitive map for object–place associations. Surprisingly, it is possible for such a complex memory structure to be formed in a few seconds. In this paper, we evaluate the memory retrieval of object–place associations in the hierarchical network formed by theta phase precession. The results show that multiple object–place associations can be retrieved with the initial cue of a scene input. Importantly, according to the wide-to-narrow unidirectional connections among scene units, the spatial area for object–place retrieval can be controlled by the spatial area of the initial cue input. These results indicate that the hierarchical cognitive maps have computational advantages on a spatial-area selective retrieval of multiple object–place associations. Theta phase precession dynamics is suggested as a fundamental neural mechanism of the human cognitive map.     

Physical models of neural networks

The state of art in computer modelling of neural networks with associative memory is reviewed. The available experimental data are considered on learning and memory of small neural systems, on isolated synapses and on molecular level. Computer simulations demonstrate that realistic models of neural ensembles exhibit properties which can be interpreted as image recognition, categorization, learning, prototype forming, etc. A bilayer model of associative neural network is proposed. One layer corresponds to the short-term memory, the other one to the long-term memory. Patterns are stored in terms of the synaptic strength matrix. We have studied the relaxational dynamics of neurons firing and suppression within the short-term memory layer under the influence of the long-term memory layer. The interaction among the layers has found to create a number of novel stable states which are not the learning patterns. These synthetic patterns may consist of elements belonging to different non-intersecting learning patterns. Within the framework of a hypothesis of selective and definite coding of images in brain one can interpret the observed effect as the "idea? generating" process.     

Modeling the effect of sleep regulation on a neural mass model

In mammals, sleep is categorized by two main sleep stages, rapid eye movement (REM) and non-REM (NREM) sleep that are known to fulfill different functional roles, the most notable being the consolidation of memory. While REM sleep is characterized by brain activity similar to wakefulness, the EEG activity changes drastically with the emergence of K-complexes, sleep spindles and slow oscillations during NREM sleep. These changes are regulated by circadian and ultradian rhythms, which emerge from an intricate interplay between multiple neuronal populations in the brainstem, forebrain and hypothalamus and the resulting varying levels of neuromodulators. Recently, there has been progress in the understanding of those rhythms both from a physiological as well as theoretical perspective. However, how these neuromodulators affect the generation of the different EEG patterns and their temporal dynamics is poorly understood. Here, we build upon previous work on a neural mass model of the sleeping cortex and investigate the effect of those neuromodulators on the dynamics of the cortex and the corresponding transition between wakefulness and the different sleep stages. We show that our simplified model is sufficient to generate the essential features of human EEG over a full day. This approach builds a bridge between sleep regulatory networks and EEG generating neural mass models and provides a valuable tool for model validation.     

主动联想记忆神经网络模型

在信息编码能提高联想记忆的存贮能力和脑内存在主动活动机制的启发下,提出一个主动联想记忆模型。模型包括两个神经网络,其一为输入和输出网络,另一个为在学习时期能自主产生兴奋模式的主动网络。两个网络的神经元之间有突触联系。由于自主产生的兴奋模式与输入无关,并可能接近于相互正交,因此,本模型有较高的存贮能力。初步分析和计算机仿真证明：本模型确有比通常联想记忆模型高的存贮能力,特别是在输入模式间有高度相关情况下、最后,对提出的模型与双向自联想记忆和光学全息存贮机制的关系作了讨论。     

Bridging behavior and physiology: ion-channel perspective on mushroom body-dependent olfactory learning and memory in Drosophila

An important body of evidence documents the differential expression of ion channels in brains, suggesting they are essential to endow particular brain structures with specific physiological properties. Because of their role in correlating inputs and outputs in neurons, modulation of voltage-dependent ion channels (VDICs) can profoundly change neuronal network dynamics and performance, and may represent a fundamental mechanism for behavioral plasticity, one that has received less attention in learning and memory studies. Revisiting three paradigmatic mutations altering olfactory learning and memory in Drosophila (dunce, leonardo, amnesiac) a link was established between each mutation and the operation of VDICs in Kenyon cells, the intrinsic neurons of the mushroom bodies (MBs). In Drosophila, MBs are essential to the emergence of olfactory associative learning and retention. Abnormal ion channel operation might underlie failures in neuronal physiology, and be crucial to understand the abnormal associative learning and retention phenotypes the mutants display. We also discuss the only case in which a mutation in an ion channel gene (shaker) has been directly linked to olfactory learning deficits. We analyze such evidence in light of recent discoveries indicating an unusual ion current profile in shaker mutant MB intrinsic neurons. We anticipate that further studies of acquisition and retention mutants will further confirm a link between such mutations and malfunction of specific ion channel mechanisms in brain structures implicated in learning and memory.     

条件联想神经网络研究

在人脑的某些功能和神经系统中的突前抑制机制启发下,本文提出一个新型的神经网络模型——条件联想神经网络.模型是一个有突触前抑制的联想记忆神经网络.通过初步分析和计算机模拟,证明本模型具有一般联想记忆模型所未有的一些新的特性,如可以在不同条件下,对同一输入有不同的反应.对同一输入,在不同的条件下,又可以有相同的反应.这些特点将有助于人们对神经系统中信息处理过程的了解.此外,文中也指出可能实现本模型的神经结构.     

A hybrid neural network of addressable and content-addressable memory

We investigate the memory structure and retrieval of the brain and propose a hybrid neural network of addressable and content-addressable memory which is a special database model and can memorize and retrieve any piece of information (a binary pattern) both addressably and content-addressably. The architecture of this hybrid neural network is hierarchical and takes the form of a tree of slabs which consist of binary neurons with the same array. Simplex memory neural networks are considered as the slabs of basic memory units, being distributed on the terminal vertexes of the tree. It is shown by theoretical analysis that the hybrid neural network is able to be constructed with Hebbian and competitive learning rules, and some other important characteristics of its learning and memory behavior are also consistent with those of the brain. Moreover, we demonstrate the hybrid neural network on a set of ten binary numeral patters     

An associative memory that can form hypotheses: a phase-coded neural network

 Nonlinear associative memories as realized, e.g., by Hopfield nets are characterized by attractor-type dynamics. When fed with a starting pattern, they converge to exactly one of the stored patterns which is supposed to be most similar. These systems cannot render hypotheses of classification, i.e., render several possible answers to a given classification problem. Inspired by von der Malsburg’s correlation theory of brain function, we extend conventional neural network architectures by introducing additional dynamical variables. Assuming an oscillatory time structure of neural firing, i.e., the existence of neural clocks, we assign a so-called phase to each formal neuron. The phases explicitly describe detailed correlations of neural activities neglected in conventional neural network architectures. Implementing this extension into a simple self-organizing network based on a feature map, we present an associative memory that actually is capable of forming hypotheses of classification. Received: 6 December 1993/Accepted in revised form: 14 July 1994     

The Wisdom of Networks: A General Adaptation and Learning Mechanism of Complex Systems

I hypothesize that re‐occurring prior experience of complex systems mobilizes a fast response, whose attractor is encoded by their strongly connected network core. In contrast, responses to novel stimuli are often slow and require the weakly connected network periphery. Upon repeated stimulus, peripheral network nodes remodel the network core that encodes the attractor of the new response. This “core‐periphery learning” theory reviews and generalizes the heretofore fragmented knowledge on attractor formation by neural networks, periphery‐driven innovation, and a number of recent reports on the adaptation of protein, neuronal, and social networks. The core‐periphery learning theory may increase our understanding of signaling, memory formation, information encoding and decision‐making processes. Moreover, the power of network periphery‐related “wisdom of crowds” inventing creative, novel responses indicates that deliberative democracy is a slow yet efficient learning strategy developed as the success of a billion‐year evolution. Also see the video abstract here: https://youtu.be/IIjP7zWGjVE .     

Neuronal Mechanisms Encoding Global-to-Fine Information in Inferior-Temporal Cortex

Sugase et al. found that global information is represented at the initial transient firing of a single face-responsive neuron in inferior-temporal (IT) cortex, and that finer information is represented at the subsequent sustained firing. A feed-forward model and an attractor network are conceivable models to reproduce this dynamics. The attractor network, specifically an associative memory model, is employed to elucidate the neuronal mechanisms producing the dynamics. The results obtained by computer simulations show that a state of neuronal population initially approaches to a mean state of similar memory patterns, and that it finally converges to a memory pattern. This dynamics qualitatively coincides with that of face-responsive neurons. The dynamics of a single neuron in the model also coincides with that of a single face-responsive neuron. Furthermore, we propose two physiological experiments and predict the results from our model. Both predicted results are not explainable by the feed-forward model. Therefore, if the results obtained by actual physiological experiments coincide with our predicted results, the attractor network might be the neuronal mechanisms producing the dynamics of face-responsive neurons.