arXiv subjects

Toshio Aoyagi

Publications and source records attributed to Toshio Aoyagi.

25 records · Page 2Linked to original sources

arXiv2008

Recurrent infomax generates cell assemblies, avalanches, and simple cell-like selectivity

Through evolution, animals have acquired central nervous systems (CNSs), which are extremely efficient information processing devices that improve an animal's adaptability to various environments. It has been proposed that the process of information maximization (infomax), which maximizes the information transmission from the input to the output of a feedforward network, may provide an explanation of the stimulus selectivity of neurons in CNSs. However, CNSs contain not only feedforward but also recurrent synaptic connections, and little is known about information retention over time in such recurrent networks. Here, we propose a learning algorithm based on infomax in a recurrent network, which we call "recurrent infomax" (RI). RI maximizes information retention and thereby minimizes information loss in a network. We find that feeding in external inputs consisting of information obtained from photographs of natural scenes into an RI-based model of a recurrent network results in the appearance of Gabor-like selectivity quite similar tothat existing in simple cells of the primary visual cortex (V1). More importantly, we find that without external input, this network exhibits cell assembly-like and synfire chain-like spontaneous activity and a critical neuronal avalanche. RI provides a simple framework to explain a wide range of phenomena observed in in vivo and in vitro neuronal networks, and it should provide a novel understanding of experimental results for multineuronal activity and plasticity from an information-theoretic point of view.

q-bio.NC↗

arXiv2007

Scale-free networks with self-growing weight

We present a novel type of weighted scale-free network model, in which the weight grows independently of the attachment of new nodes. The evolution of this network is thus determined not only by the preferential attachment of new nodes to existing nodes but also by self-growing weight of existing links based on a simple weight-driven rule. This model is analytically tractable, so that the various statistical properties, such as the distribution of weight, can be derived. Finally, we found that some type of social networks is well described by this model.

cond-mat.dis-nn↗

arXiv2004

Effect of Synchronous Incoming Spikes on Activity Pattern in A Network of Spiking Neurons

Although recent neurophysiological experiments suggest that synchronous neural activity is involved in some perceptual and cognitive processes, the functional role of such coherent neuronal behavior is not well understood. As a first step in clarifying this role, we investigate how the temporal coherence of certain neuronal activity affects the activity pattern in a neural network. Using a simple network of leaky integrate-and-fire neurons, we study the effects of synchronized incoming spikes on the functioning of two mechanisms typically used in model neural systems, winner-take-all competition and associative memory. We demonstrate that a pair of switches undergone by the incoming spikes, from asynchronous to synchronous and then back to asynchronous, triggers a transition of the network from one state to another state. In the case of associative memory, for example, this switching controls the timing of the next recalling, whereas the firing rate pattern in the asynchronous state prepares the network for the next retrieval pattern.

q-bio.NC↗

arXiv2000

Analysis of Oscillator Neural Networks for Sparsely Coded Phase Patterns

We study a simple extended model of oscillator neural networks capable of storing sparsely coded phase patterns, in which information is encoded both in the mean firing rate and in the timing of spikes. Applying the methods of statistical neurodynamics to our model, we theoretically investigate the model's associative memory capability by evaluating its maximum storage capacities and deriving its basins of attraction. It is shown that, as in the Hopfield model, the storage capacity diverges as the activity level decreases. We consider various practically and theoretically important cases. For example, it is revealed that a dynamically adjusted threshold mechanism enhances the retrieval ability of the associative memory. It is also found that, under suitable conditions, the network can recall patterns even in the case that patterns with different activity levels are stored at the same time. In addition, we examine the robustness with respect to damage of the synaptic connections. The validity of these theoretical results is confirmed by reasonable agreement with numerical simulations.

cond-mat.dis-nn↗

arXiv1998

An Oscillator Neural Network Retrieving Sparsely Coded Phase Patterns

Little is known theoretically about the associative memory capabilities of neural networks in which information is encoded not only in the mean firing rate but also in the timing of firings. Particularly, in the case that the fraction of active neurons involved in memorizing patterns becomes small, it is biologically important to consider the timings of firings and to study how such consideration influences storage capacities and quality of recalled patterns. For this purpose, we propose a simple extended model of oscillator neural networks to allow for expression of non-firing state. %which is able to memorize sparsely coded phase patterns including non-firing states. Analyzing both equilibrium states and dynamical properties in recalling processes, we find that the system possesses good associative memory.

cond-mat.dis-nn↗

arXiv1998

Retrieval dynamics of neural networks for sparsely coded sequential patterns

It is well known that a sparsely coded network in which the activity level is extremely low has intriguing equilibrium properties. In the present work, we study the dynamical properties of a neural network designed to store sparsely coded sequential patterns rather than static ones. Applying the theory of statistical neurodynamics, we derive the dynamical equations governing the retrieval process which are described by some macroscopic order parameters such as the overlap. It is found that our theory provides good predictions for the storage capacity and the basin of attraction obtained through numerical simulations. The results indicate that the nature of the basin of attraction depends on the methods of activity control employed. Furthermore, it is found that robustness against random synaptic dilution slightly deteriorates with the degree of sparseness.

cond-mat.dis-nn↗

arXiv1994

A Network of Oscillators for Retrieving Phase Information

We propose a network of oscillators to retrieve given patterns in which the oscillators keep a fixed phase relationship with one another. In this description, the phase and the amplitude of the oscillators can be regarded as the timing and the strength of the neuronal spikes, respectively. Using the amplitudes for encoding, we enable the network to realize not only oscillatory states but also non-firing states. In addition, it is shown that under suitable conditions the system has a Lyapunov function ensuring a stable retrieval process. Finally, the associative memory capability of the network is demonstrated numerically.

adap-org↗