Biological agents navigate complex environments by combining long-term memory of successful actions with short-term suppression of recently visited locations-a capability that remains difficult to replicate in artificial systems, especially under partial observability. Inspired by the complementary timescales of neural and astrocytic dynamics, we introduce a spiking neuron-astrocyte network (SNAN) where spike-timing-dependent plasticity (STDP) reinforces successful action sequences on a distant time scale, while astrocytic calcium transients suppress recently visited states on a short-term time scale, effectively blocking locations already explored. This dual-timescale memory mechanism biases the agent toward unexplored regions, accelerating goal finding without requiring explicit global statistics. We show that in grid-world navigation tasks with extreme partial observability, SNAN reduces median path length by up to sixfold and drastically improves goal completion rates compared to baseline agents. The astrocytic modulation inherently mitigates the exploration-exploitation trade-off as an emergent consequence of local state suppression. This kind of local sensory data modulation can be considered as a new type of working memory referred to as a "Topological-Context Memory". To validate hardware feasibility using neuromorphic approaches, we map STDP to a memristive VTEAM model and implement a subset of the network on a crossbar array, achieving order-of-magnitude gains in speed per area and energy per decision over CPU implementations. Our results establish astrocyte-inspired dual-timescale memory as a scalable, hardware-realizable principle for neuromorphic robotics and edge-AI systems.

Taste stimuli are encoded in the brain through distributed spatiotemporal activity patterns, which motivates their description as continuous multidimensional variables rather than discrete perceptual categories. In this work, we introduce TASTE-SNN—a biologically plausible spiking neural network designed to decode four-dimensional taste stimuli representing sweetness, sourness, saltiness, and bitterness. The proposed architecture integrates Gaussian population coding, adaptive leaky integrate-and-fire neurons, competitive inhibition, memristive inhibitory spike-timing-dependent plasticity (STDP), and astrocyte-like modulation of glutamatergic synapses into a compact neuromorphic framework. Hidden-layer activity is mapped to reconstructed taste vectors via a Ridge regression decoder. Training combines unsupervised winner-take-all STDP in the first layer with curriculum-based optimization of deeper synapses using a local three-factor learning rule. The resulting network exhibits stable firing-rate regulation, bounded synaptic evolution, and progressively decreasing validation loss across curriculum stages. With only 1928 trainable parameters, TASTE-SNN provides a compact and interpretable model for studying biologically grounded sensory decoding, especially in resource-constrained or neuromorphic implementations.

Spiking neural networks (SNNs) trained with spike-timing-dependent plasticity (STDP) provide a biologically plausible and energy-efficient framework for unsupervised pattern recognition, yet their accuracy remains substantially below that of supervised deep networks. In this work, we incorporate the extracellular-matrix (ECM) neuron activity regulatory model into an STDP-based SNN. The ECM introduces four dynamical variables (Q, Z, P, R) coupled to each excitatory neuron via two mechanisms: an activity-dependent additive threshold term and a homeostatic synaptic scaling term. On the 400-neuron benchmark trained with 10,000 MNIST samples and evaluated on the 10,000-sample test set, the full ECM model achieved 86.77

Development of biomorphic underwater robots employing locomotion principles of aquatic organisms represents an actively advancing area of modern robotics. Of particular interest are robots implementing thunniform locomotion, characteristic of highly efficient swimmers. This paper presents a fish-like robot whose design, based on the morphology of the yellowfin tuna, is optimized for experimental studies of swimming kinematics under laboratory conditions. The robot is equipped with a caudal propulsion system that generates forward motion through periodic oscillations of the tail fin, mimicking natural swimming kinematics. The article describes the robot’s structural features, internal layout, sensors, and principles of three-dimensional maneuvering. Experimental swimming kinematics studies were conducted in a pool using video recording followed by computer vision data processing. Swimming speed was determined under different operating modes of the tail actuator based on marker coordinates on the robot’s body. The dependence of swimming speed and energy efficiency on tail beat frequency at constant amplitude was investigated. Swimming speed increases approximately linearly with increasing frequency. The minimum Cost of Transport (COT) value was 6.22 J·kg⁻¹·m⁻¹, which is comparable to biological prototypes and the most efficient biomorphic robots. These results can inform the further design of energy-efficient fish-like underwater vehicles

Live neural systems demonstrate remarkable capabilities to learn new behavior and patterns from mere few examples and are known to operate robustly under severe sensory noise. These capabilities, however, remain largely out of reach for modern artificial neural networks, including deep learning models. We show that this gap can be bridged by embedding novel genuine neuromorphic circuits into conventional artificial neural network architectures. These circuits comprise astrocytic modulation and spiking dynamics inherent to biological neural structures. Tested across standard benchmarks representing tasks of varying complexity, the hybrid models achieve high accuracy from few training examples per class and sustain high performance under occlusion and impulse noise that cause performance collapse in standard models without neuromorphic adaptation. We term this phenomenon neuromorphic supremacy - a regime in which architectures grounded in neurobiology decisively outperform classical deep learning, pointing toward a principled foundation for perception in embodied AI systems operating in noisy, data-scarce environments.

A nonlinear mathematical model describing the vertical motion of a biomimetic underwater vehicle equipped with a swim bladder is developed. For a passive bladder that changes its volume under hydrostatic pressure, the system can achieve depth stabilization through a speed-induced mechanism when the lever-arm geometry (relative positions of the swim bladder and lifting surfaces) is favorable; otherwise stabilization is not possible. Using the Routh–Hurwitz criterion, analytical stability conditions are obtained in closed form, revealing a lower onset speed that is set by a simple coupling between forward speed and geometry. Numerical simulations confirm the theoretical predictions and reveal the dominant loss-of-stability scenarios: loss of effective stiffness at the onset threshold and oscillatory instability when the mixed speed–geometry factor changes sign. The results demonstrate the feasibility of passive swim-bladder-based depth stabilization and provide practical guidelines for selecting the lever arms of the buoyancy and lift forces and operating speeds in autonomous underwater vehicles.

In this study, we address the issue of whether applying transcranial magnetic and vibrotactile stimulation can improve the motor imagery BMI performance. Our findings provide evidence that applying transcranial magnetic stimulation with specified parameters (frequency 5 Hz, duration 6 min, 90

The purpose of this work is to study the role of mechanisms of astrocytic regulation of synaptic transmission in the processes of synchronization formation in signaling of neural networks by mathematical modeling methods. Methods. The paper presents a model of a small neuron-astrocyte ensemble. The HodgkinHuxley model is used as a model of the membrane potential dynamics of a neuron. The case of an ordered topology of connections ("all-to-all") in a neural network is considered. The astrocyte network is modeled as a network of diffusion-coupled calcium oscillators with an ordered topology (in which the matrix of connections is structured in a certain way, interaction with the nearest neighbors). A biophysical model of calcium dynamics is used as an astrocyte model. The effect of astrocytes on neurons is taken into account as a slow modulation of synaptic connections weights in the neural network, proportional to calcium signals in nearby astrocytes. In other words, at the network level, the possibility of adaptive restructuring of oscillatory wave patterns due to astrocyte-induced regulation of synaptic transmission is being studied. The synchronization of neuronal activity is estimated by calculating the coherence of the neural network signaling. Results. The influence of astrocytes on the dynamics of the neural network consists in the excitation of time-correlated patterns of neural activity caused by an astrocytedependent increase in synaptic interaction between neurons on the time scales of astrocytic dynamics. It has been shown that synchronized calcium signaling of the astrocytic network leads to coordinated burst (bundle) activity of the neural network, which occurs against the background of uncorrelated spontaneous impulse activity induced by external noise stimulation. The influence of specific biophysical mechanisms of astrocytic modulation of synaptic transmission on the dynamic properties of local synchronization structures in neural ensembles has been investigated. The characteristics of the coordinated bundle activity of a neural network are studied depending on the properties of external noise stimulation, the strength of astrocytic regulation of synaptic transmission, as well as the degree of neurons influence on astrocytes.

Here we present a closed-loop control system for real-time maintenance of hippocampal neuronal activity, integrating artificial neural networks with memristive technologies. The system performs detection of local field potentials (LFP) followed by automatic assessment of their amplitude. Adaptive stimulation, whose parameters are adjusted by a neural network algorithm via a memristive interface, minimizes the difference between current and target LFP amplitudes, ensuring proportional stimulus correction. The developed approach demonstrates a new standard of efficacy and safety for therapeutic interventions through its ability to dynamically adapt stimulation parameters to the changing state of neural networks.

Mammalian brains operate in very special surroundings: to survive they have to react quickly and effectively to the pool of stimuli patterns previously recognized as danger. Many learning tasks often encountered by living organisms involve a specific set-up centered around a relatively small set of patterns presented in a particular environment. For example, at a party, people recognize friends immediately, without deep analysis, just by seeing a fragment of their clothes. This set-up with reduced "ontology" is referred to as a "situation." Situations are usually local in space and time. In this work, we propose that neuron-astrocyte networks provide a network topology that is effectively adapted to accommodate situation-based memory. In order to illustrate this, we numerically simulate and analyze a well-established model of a neuron-astrocyte network, which is subjected to stimuli conforming to the situation-driven environment. Three pools of stimuli patterns are considered: external patterns, patterns from the situation associative pool regularly presented to the network and learned by the network, and patterns already learned and remembered by astrocytes. Patterns from the external world are added to and removed from the associative pool. Then, we show that astrocytes are structurally necessary for an effective function in such a learning and testing set-up. To demonstrate this we present a novel neuromorphic computational model for short-term memory implemented by a two-net spiking neural-astrocytic network. Our results show that such a system tested on synthesized data with selective astrocyte-induced modulation of neuronal activity provides an enhancement of retrieval quality in comparison to standard spiking neural networks trained via Hebbian plasticity only. We argue that the proposed set-up may offer a new way to analyze, model, and understand neuromorphic artificial intelligence systems.

The formation of functional connections between cells derived from neuronal progenitor cells (NPCs) and the developed neuronal network is a crucial task in neurobiology and cellular technologies for medical applications. We developed a technique to co-culture NPCs and mature neuronal cells in a three-chamber microfluidic chip in vitro and studied the features of functional connections among cells derived from NPCs during their integration into a network of differentiated neurons. Starting from day 20 of primary culture development, a network forms from integrated NPCs, which ensures the distribution of bioelectrical activity among neurons in a manner similar to that of differentiated cells.

This paper proposes a theoretical framework for optimizing lift force generation in flapping-wing systems combining analytical and data-driven methods to identify optimal parameters for wing oscillations. A theoretical optimization model for lifting force generation by oscillating wings was developed. The model employed a two-dimensional analysis of the interaction between the flapping wing and the surrounding airflow. An analytical expression for the lift force was derived, and the dependencies of lift on both the angle of attack and wing deflection angle were investigated. Our analysis revealed the existence of an optimal angle of attack that maximizes lift, a finding that presents a crucial control parameter for implementing intellectual, performance-maximizing flight maneuvers. The theoretical results were validated through experimental measurements of the lift force produced by a commercially available flapping-wing ornithopter.

The purpose of this work is to study the effects of spatio-temporal dynamics of spontaneous calcium signaling in the morphological structure of an astrocyte at the subcellular level using biophysical mathematical modeling methods. Methods. This work proposes a biophysical multicompartmental model of noise-induced calcium dynamics in the astrocytic process. The model describes the process of generation of spontaneous Ca2+ signals induced by the stochastic activation of voltage-dependent Ca2+ channels on the plasma membrane of the astrocyte. The model allows us to study the dynamics of the propagation of spontaneous local Ca2+ signals and the mechanisms of formation of spatial Ca2+ patterns in the astrocytic process. Results. The developed model enables studying the influence of morphology and intracellular biophysical mechanisms on the characteristics of spontaneous noise-induced Ca2+ signaling in the astrocytic process. The parameter ranges at which the model qualitatively reproduces the spontaneous Ca2+ activity at the subcellular level observed in experimental studies have been specified. The characteristics of noise-induced Ca2+ patterns propagating along the process were investigated, depending on the internal structure of the process, its geometry, and the steady state concentration of inositol 1,4,5-triphosphate molecules.

Memory consolidation in the brain involves complex processes at both synaptic and systems levels, with synaptic plasticity playing a key role in stabilizing recently encoded memories. However, the brain’s ability to forget is as crucial as its ability to remember, raising the plasticity-stability dilemma. This study explores the role of structural plasticity, specifically synaptic rewiring driven by spike-timing-dependent plasticity (STDP), in memory reliability and forgetting within spiking neural networks (SNNs). Using a computational model based on the Izhikevich spiking neuron dynamics, we investigate how STDP-driven rewiring influences the stability of memory traces under spontaneous activity. Our results demonstrate that rewiring significantly enhances memory reliability, allowing networks to retain information for extended periods compared to solely synaptic plasticity action. Noticeably, that the consolidation of memory traces induces specific changes in anatomic connectome, facilitated by rewiring during spontaneous activity. The emergence of hub neurons, capable of triggering population bursts, further supports the network’s ability to retain information. These findings suggest that rewiring, alongside synaptic plasticity, plays a crucial role in balancing learning and forgetting, offering insights into the mechanisms underlying memory consolidation in biologic neural networks.

In this manuscript, we investigate the memristor-based implementation of neuronal ion channels in a mathematical model and an experimental circuit for a neuronal oscillator. We used a FitzHugh-Nagumo equation system describing neuronal excitability. Non-linearities introduced by the voltage-gated ion channels were modeled using memristive devices. We implemented three basic neuronal excitability modes including the excitable mode corresponding to a single spike generation, self-oscillation stable limit cycle mode with periodic spike trains and bistability between a fixed point and a limit cycle. We also found the spike-burst activity of mathematical and experimental models under certain system parameters. Modeling synaptic transmission, we simulated postsynaptic response triggered by periodic pulse stimulation. We found that due to the charge accumulation effect in the memristive device, the electronic synapse implemented a qualitatively bio-plausible synapse with a potentiation effect with increasing amplitude of the response triggered by a spike sequence.

The use of central rhythm generators (CPGs) as control systems has recently received more and more attention. Such systems allow us to solve the problem of path prediction. They have higher stability. However, there are a number of problems in designing successful and working systems such as a real robot -CPG. For example, it is necessary to synchronize the operating time of the CPG model and the actuator. There are also difficulties with using CPG on different robots. Even when modeling robots with similar prototypes: fish, salamanders, snakes, it is necessary not only to change the number of segments of the control system to control the actuators, but also to completely change the operating parameters of the CPG for proper interaction with the environment. Next, we demonstrate a CPG capable of controlling a virtual robot fish and a ground robot snake. The model is capable of moving forward and turning left and right. It is possible to connect external sensors to transmit information about an obstacle to the system and perform a turn under its influence.

The article describes the design and main dynamic characteristics of an underwater vehicle of the biomorphic (fish-like) type, which implements the tunniform type of locomotion for movement under water. The body of the device is made on the basis of a digital model of the body of yellowfin tuna. The movement is provided by a driv-ing device that simulates the operation of the tail fin of a fish. The robot's navigation on the surface and with a dive to a shallow depth is controlled by a remote control system, which allows you to change the dynamic characteris-tics (amplitude and frequency) of tail vibrations. In the experimental study, the dependencies of the speed of movement and energy consumption of the robot depending on these characteristics were obtained. In the theoreti-cal study, a computational model of the robot was obtained based on the solution of hydrodynamic equations. In computer modeling of swimming using the method of deformable nets, a good correspondence with experimental data was obtained. In addition, the hydrodynamic characteristics of the flow during the movement of the robot were investigated, as well as various features of the vibrations of the elements of the robot body when moving in a liquid were explained.

In the context of natural disasters, human responses inevitably intertwine with natural factors. The COVID-19 pandemic, as a significant stress factor, has brought to light profound variations among different countries in terms of their adaptive dynamics in addressing the spread of infection outbreaks across different regions. This emphasizes the crucial role of cultural characteristics in natural disaster analysis. The theoretical understanding of large-scale epidemics primarily relies on mean-field kinetic models. However, conventional SIR-like models failed to fully explain the observed phenomena at the onset of the COVID-19 outbreak. These phenomena included the unexpected cessation of exponential growth, the reaching of plateaus, and the occurrence of multi-wave dynamics. In situations where an outbreak of a highly virulent and unfamiliar infection arises, it becomes crucial to respond swiftly at a non-medical level to mitigate the negative socio-economic impact. Here we present a theoretical examination of the first wave of the epidemic based on a simple SIRSS model (SIR with Social Stress). We conduct an analysis of the socio-cultural features of naïve population behaviors across various countries worldwide. The unique characteristics of each country/territory are encapsulated in only a few constants within our model, derived from the fitted COVID-19 statistics. These constants also reflect the societal response dynamics to the external stress factor, underscoring the importance of studying the mutual behavior of humanity and natural factors during global social disasters. Based on these distinctive characteristics of specific regions, local authorities can optimize their strategies to effectively combat epidemics until vaccines are developed.

In this article, we propose a novel model for the regulation of neuronal network activity during astrocytic modulation of synaptic plasticity. The work also examines the influence of astrocytic modulation of STDP on the sensitivity of synapses to spatiotemporal patterns.

This study presents a novel mathematical model of astrocytic regulation of the dynamics of central pattern generators. The neuron model is based on the Komendantov-Kononenko model. Astrocytic dynamics are described by the mean-field approach. The model considers homeostatic astrocytic regulation of neuronal activity.