Neurons encode information about the environment through their activity. As animals explore the environment, neurons rapidly acquire selectivity for distinct features of the external world; characterizing how these selectivity patterns emerge, reorganize, and overlap is key to linking neural activity to behavior and cognition. Calcium imaging in freely behaving animals can record large neuronal populations, but quantifying neuron-behavior selectivity directly from continuous fluorescence is challenging because both signals are temporally autocorrelated and calcium kinetics introduce time lags. Here we present INTENSE (INformation-Theoretic Evaluation of Neuronal SElectivity), an open-source framework that uses mutual information to detect neuron-behavior associations from raw calcium fluorescence data. INTENSE controls false discoveries using circular-shift permutation testing that preserves temporal structure and optimizes temporal delays to account for indicator kinetics and prospective/retrospective encoding. To separate genuine mixed selectivity from associations driven by behavioral covariance, INTENSE applies conditional mutual information-based disentanglement. We validated INTENSE on synthetic datasets, demonstrating robust detection across diverse signal-to-noise ratios and reliability conditions, whereas methods lacking temporal controls show poor performance. Applied to CA1 miniscope recordings in mice freely exploring an open field, INTENSE reveals robust selectivity to multiple variables (place, head direction, object interaction, locomotion) and refines mixed-selectivity estimates by distinguishing redundant from genuinely multi-variable encoding. Together, INTENSE enables high-throughput, information-theoretic selectivity mapping with principled control of temporal structure and behavioral covariance, bridging large-scale recordings to circuit-level hypotheses.

Brain activity spans single-neuron, population, and network levels, and core questions in neural coding require moving between them. Yet current tools target a single paradigm and incompatible data formats, leaving cross-level questions hard to address. We present DRIADA, an open-source Python framework that unifies neural signals and time-aligned behavior in a shared data model, so selectivity testing, dimensionality reduction, and network analysis operate within a unified workflow. We evaluate it on synthetic data with known ground truth, hippocampal calcium imaging from 13 mice in an open field, and a simulated toroidal attractor network. In the hippocampal data, selectivity-based filtering restored a two-dimensional spatial embedding from a collapsed all-neuron embedding, while reverse analysis showed that ∼57% of neurons informative about leading manifold dimensions were not selective to any of the 11 measured behavioral features. On the toroidal benchmark, four independent modules recovered the expected topology. DRIADA makes cross-scale analysis routine across calcium imaging, spike trains, and simulated networks.

Understanding spatial navigation critically depends on uncovering how animals encode environmental information. Place cells in the hippocampus are essential for constructing spatial maps by encoding targets, landmarks, boundaries, and other environmental features. However, it has remained unclear whether or not specific mapping properties emerge in respect to contrasting environmental elements that do not physically restrict animal movement. In this study, we investigated the calcium activity of CA1 hippocampal neurons in mice using miniscope imaging while the animals explored an arena with contrasting colored elements incorporated into the floor design. Our results show that the place fields of CA1 neurons are preferentially concentrated in regions of environmental heterogeneity rather than within the contrasting elements themselves or in zones preferred by the mice. These findings suggest that CA1 hippocampal neurons integrate surface contrast signals into the broader spatial representation of the environment.

A detailed analysis of animal behavior is a crucial step in understanding how the activity of brain structures and individual cells mediate animal behavior. Recent advances in spatial navigation research highlight the need for 3D behavioral paradigms, as traditional 2D mazes cannot fully capture naturalistic behavior in volumetric space. Here we present an extension of the Sphynx software package that enables robust animal 3D trajectory reconstruction. We have shown possibility to extract a wide range of behavioral variables from video recordings of animal behavior in three-dimensional mazes.

Elements of neural networks, both biological and artificial, can be described by their selectivity for specific cognitive features. Understanding these features is important for understanding the inner workings of neural networks. For a living system, such as a neuron, whose response to a stimulus is unknown and not differentiable, the only way to reveal these features is through a feedback loop that exposes it to a large set of different stimuli. The properties of these stimuli should be varied iteratively in order to maximize the neuronal response. To utilize this feedback loop for a biological neural network, it is important to run it quickly and efficiently in order to reach the stimuli that maximizes certain neurons' activation with the least number of iterations possible. Here we present a framework with an efficient design for such a loop. We successfully tested it on an artificial spiking neural network (SNN), which is a model that simulates the asynchronous spiking activity of neurons in living brains. Our optimization method for activation maximization is based on the low-rank Tensor Train decomposition of the discrete activation function. The optimization space is the latent parameter space of images generated by SN-GAN or VQ-VAE generative models. To our knowledge, this is the first time that effective AM has been applied to SNNs. We track changes in the optimal stimuli for artificial neurons during training and show that highly selective neurons can form already in the early epochs of training and in the early layers of a convolutional spiking network. This formation of refined optimal stimuli is associated with an increase in classification accuracy. Some neurons, especially in the deeper layers, may gradually change the concepts they are selective for during learning, potentially explaining their importance for model performance.

The manifold hypothesis in application to neural data states that brain activity patterns are confined to a low-dimensional manifold that spans a small fraction of the theoretically available state space. We test this hypothesis using in vivo calcium imaging data from hippocampal neurons in mice during novel arena exploration. We compute the intrinsic dimensionality of neural activity and compare it with other dimensionality estimates. Due to the inevitable scarcity of in vivo recordings, both in terms of neurons recorded and duration, we follow the dimensionality patterns as the amount of available data increases. We find a nontrivial power-law scaling of intrinsic dimension with the number of recorded neurons, and discuss its potential biological significance.

Encoding environmental information is a primary focus in neuroscience, as environmental context is crucial for the spatial orientation of animals. Place cells in the hippocampus play a key role in forming the spatial map of the surrounding space by mapping targets, environmental objects, various landmarks, and boundaries. However, it was previously unclear whether specific mapping properties exist for contrasting environmental elements that do not physically obstruct animal movement. In this study, we examined the calcium activity of CA1 hippocampal neurons in mice using miniscope imaging, alongside their behavior in an arena with contrasting colored elements on the floor. Our findings indicate that the place fields of CA1 hippocampal cells are concentrated in areas of environmental heterogeneity, rather than in the individual elements themselves or in areas preferred by the mice. This suggests that the contrasting surface signals of the environment are represented by CA1 hippocampal neurons within the overall spatial pattern.

A detailed analysis of animal behavior is a crucial step in understanding how the activity of brain structures and individual cells mediate animal behavior. Neurons can be selective to both the internal states of the animal and features of the environment. However, it is challenging to divide the continuum of animal behavior into discrete behavioral acts. We present Sphynx, a software tool for extracting a wide range of behavioral variables from video recordings of animal behavior and correlating these acts with environmental features. We demonstrate that detailed behavioral analysis using Sphynx reveals selectivity of neurons.

In this study, we explore the fundamental principles behind the architecture of the human brain's structural connectome, from the perspective of spectral analysis of Laplacian and adjacency matrices. Building on the idea that the brain strikes a balance between efficient information processing and minimizing wiring costs, we aim to understand the impact of the metric properties of the connectome and how they relate to the existence of an inherent scale. We demonstrate that a simple generative model, combining nonlinear preferential attachment with an exponential penalty for spatial distance between nodes, can effectively reproduce several key characteristics of the human connectome, including spectral density, edge length distribution, eigenmode localization and local clustering properties. We also delve into the finer spectral properties of the human structural connectomes by evaluating the inverse participation ratios (IPR_q) across various parts of the spectrum. Our analysis reveals that the level statistics in the soft cluster region of the Laplacian spectrum deviate from a purely Poisson distribution due to interactions between clusters. Additionally, we identified scar-like localized modes with large IPR values in the continuum spectrum. We identify multiple fractal eigenmodes distributed across different parts of the spectrum, evaluate their fractal dimensions and find a power-law relationship in the return probability, which is a hallmark of critical behavior. We discuss the conjectures that a brain operates in the Griffiths or multifractal phases.

This study investigates the dynamics of non-spatial specializations in hippocampal place cells during exposure to novel environments. Hippocampal place cells, known for their role in spatial mapping, exhibit multi-modal responses to sensory cues. The research focuses on understanding how these cells adapt their specialization in response to novel stimuli, specifically examining non-spatial determinants such as odors and social interactions. Using a social-driven food odor recognition model in mice, the study records CA1 hippocampal neuron activity through miniscope imaging. The experimental design involves demonstrations of novel odors to mice, followed by observation sessions with food options. The analysis employs deep neural network tools for behavior tracking and the custom-developed INTENS software package for identifying neural specializations. Results indicate multiple specializations, particularly those related to odor, with differences observed between training and testing sessions. The findings suggest a temporal aspect to the formation of these specializations in novel conditions, necessitating further investigation for precise tracking.

We studied the population activity of hippocampal neurons in mice during novel arena exploration. To quantify the partial synchronization of neuronal activity, we calculated the effective dimensionality of a multidimensional time series of neuronal activity obtained using in vivo using calcium imaging. The effective dimensionality was calculated from the spectra of correlation matrices subjected to the consistent bias correction procedure. Applying this algorithm to mouse neural activity data, we found that the effective activity dimensionality of neuronal populations significantly increased during periods of animal stops. We attribute this finding to the presence of functional ensembles of neurons associated with movements, whose synchronised activity was disrupted during stops.

We investigate the statistics of the largest eigenvalue, $\lambda_{\rm max}$, in an ensemble of $N\times N$ large ($N\gg 1$) sparse adjacency matrices, $A_N$. The most attention is paid to the distribution and typical fluctuations of $\lambda_{\rm max}$ in the vicinity of the percolation threshold, $p_c=\frac{1}{N}$. The overwhelming majority of subgraphs representing $A_N$ near $p_c$ are exponentially distributed linear subchains, for which the statistics of the normalized largest eigenvalue can be analytically connected with the Gumbel distribution. For the ensemble of {\rm all} subgraphs near $p_c$ we suggest that under an appropriate modification of the normalization constant the Gumbel distribution provides a reasonably good approximation. Using numerical simulations we demonstrate that the proposed transformation of $\lambda_{\rm max}$ is indeed Gumbel-distributed and the leading finite-size corrections in the vicinity of $p_c$ scale with $N$ as $\sim \ln^{-2}N$. All together, our results reveal a previously unknown universality in eigenvalue statistics of sparse matrices close to the percolation threshold.

Hippocampal place cells are a well-known object in neuroscience, but their place field formation in the first moments of navigating in a novel environment remains an ill-defined process. To address these dynamics, we performed in vivo imaging of neuronal activity in the CA1 field of the mouse hippocampus using genetically encoded green calcium indicators, including the novel NCaMP7 and FGCaMP7, designed specifically for in vivo calcium imaging. Mice were injected with a viral vector encoding calcium sensor, head-mounted with an NVista HD miniscope, and allowed to explore a completely novel environment (circular track surrounded by visual cues) without any reinforcement stimuli, in order to avoid potential interference from reward-related behavior. First, we calculated the average time required for each CA1 cell to acquire its place field. We found that 25% of CA1 place fields were formed at the first arrival in the corresponding place, while the average tuning latency for all place fields in a novel environment equaled 247 s. After 24 h, when the environment was familiar to the animals, place fields formed faster, independent of retention of cognitive maps during this session. No cumulation of selectivity score was observed between these two sessions. Using dimensionality reduction, we demonstrated that the population activity of rapidly tuned CA1 place cells allowed the reconstruction of the geometry of the navigated circular maze; the distribution of reconstruction error between the mice was consistent with the distribution of the average place field selectivity score in them. Our data thus show that neuronal activity recorded with genetically encoded calcium sensors revealed fast behavior-dependent plasticity in the mouse hippocampus, resulting in the rapid formation of place fields and population activity that allowed the reconstruction of the geometry of the navigated maze.

A resting state network is a correlated activity of many neural structures in absence of external stimulation or functional tasks, and it is a fundamental endogenous feature of the human and animal brain. However, the nature and function of such spontaneous activity remain poorly understood. One of the hypothesis suggest that they reflect background replay and consolidation of individually acquired neural networks of prior experience. However, classical non-invasive methods used for resting state network detection cannot tag specific cellular elements of the neural network at the time when individual experience is acquired so that their activity can be subsequently investigated in the resting state. To overcome this limitation, we started a project on cellular imaging of mouse resting state networks and relating their activity to animal's past experiences. We used a large-scale c-Fos imaging of resting state neuronal activity in the mouse brain combined with graph analysis methods to get deeper insights into structure and functional significance of resting state brain networks. We characterized resting-state activity of 104 mouse brain structures and found that there was no direct relationship between anatomical attributes of examined areas and the level of their activity. We also analyzed individual variability of brain areas activity and showed that resting-state networks identified by c-Fos expression were stable and reproducible in all the animals. Next, c-Fos activity of 42 selected brain areas (sensory and motor cortices, hippocampus, parahippocampal cortex, amygdala, basal nuclei, associative and sensory thalamic nuclei, hypothalamic nuclei and midbrain) to characterize the major components and analyze functional connectivity of the resting state network. we identified several major groups of functionally connected areas in the resting state network of awake mouse brain: a cluster of medial prefrontal cortex and other medial associative neocortical areas, a cluster of visual areas, a tightly connected cluster of sensorimotor areas and basal nuclei, and a fully isolated cluster of auditory areas. Importantly, activity of structures known for their relationship to fear and threat learning (such as hippocampus, amygdala, and prelimbic cortex) was not correlated and did not comprise any functional group. This high variability in the activity of fear-related brain structures will be used at the next stage of the project to examine changes in the resting state network activity in relation to prior threat experience.

We study the problem of designing scalable algorithms to find effective intervention strategies for controlling stochastic epidemic processes on networks. This is a common problem arising in agent based models for epidemic spread. Previous approaches to this problem focus on either heuristics with no guarantees or approximation algorithms that scale only to networks corresponding to county-sized populations, typically, with less than a million nodes. In particular, the mathematical-programming based approaches need to solve the Linear Program (LP) relaxation of the problem using an LP solver, which restricts the scalability of this approach. In this work, we overcome this restriction by designing an algorithm that adapts the multiplicative weights update (MWU) framework, along with the sample average approximation (SAA) technique, to approximately solve the linear program (LP) relaxation for the problem. To scale this approach further, we provide a memory-efficient algorithm that enables scaling to large networks, corresponding to country-size populations, with over 300 million nodes and 30 billion edges. Furthermore, we show that this approach provides near-optimal solutions to the LP in practice.

We investigated the structure of the functional networks of the human brain using data from the OASIS dataset. To construct functional networks from BOLD signals, we used the method of dynamic time warping (DTW), which allows one to take into account possible distortions and nonlinear effects when comparing two time series. We investigated the resulting functional networks in terms of graph entropy, a recently proposed thermodynamic approach to describing dynamics in complex networks. The graph entropy approach provides tools to investigate the information flows in networks on different timescales. We showed a high heterogeneity of the resulting individual functional networks, expressed in a significant mismatch of the characteristic excitation diffusion times between subjects. We also constructed an artificial network model with a hierarchy of temporal scales to explain the detected multiscale nature of some functional networks. No differences were found between the healthy subjects and subjects with different levels of clinical dementia rating. We hypothesize that the heterogeneity we found may be related to the personality traits of the subjects, and we intend to investigate this issue further.

We expose a series of exact mappings between particular cases of four statistical physics models: (i) equilibrium 1D lattice gas with nearest-neighbor repulsion, (ii) (1 + 1)D combinatorial heap of pieces, (iii) directed random walks on a half-plane, and (iv) 1D totally asymmetric simple exclusion process (TASEP). In particular, we show that generating function of a 1D steady-state TASEP with open boundaries can be interpreted as a quotient of partition functions of 1D hard-core lattice gases with one adsorbing lattice site and negative fugacity. This result is based on the combination of a representation of a steady-state TASEP configurations in terms of (1 + 1)D heaps of pieces (HP) and a theorem of X Viennot which projects the partition function of (1 + 1)D HP onto that of a single layer of pieces, which in this case is a 1D hard-core lattice gas.

We demonstrate here a series of exact mappings between particular cases of four statistical physics models: equilibrium 1-dimensional lattice gas with nearestneighbor repulsion, (1 + 1)-dimensional combinatorial heap of pieces, random walks on half-plane and totally asymmetric simple exclusion process (TASEP) in one dimension (1D). In particular, we show that generating function of a steady state of one-dimensional TASEP with open boundaries can be interpreted as a quotient of partition functions of 1D hard-core lattice gases with one adsorbing lattice site and negative fugacity. This result is based on the combination of (i) a representation of the steady-state TASEP configurations in terms of (1+1)-dimensional heaps of pieces and (ii) a theorem connecting the partition function of (1 + 1)-dimensional heaps of pieces with that of a single layer of pieces, which in this case is a 1D hard-core lattice gas.

The brain at wakefulness is active even in the absence of goal-directed behavior or salient stimuli. However, patterns of this resting-state (RS) activity can undergo long-term alterations following exposure to preceding meaningful stimuli. This study was aimed to develop an unbiased method to detect such changes in the RS activity after exposure to emotionally meaningful stimuli. For this purpose, we used functional magnetic resonance imaging (fMRI) of RS brain activity before and after acquisition and extinction of experimental conditioned fear. A group of healthy volunteers participated in three fMRI sessions: a RS before fear conditioning, a fear extinction session, and a RS immediately after fear extinction. The fear-conditioning paradigm consisted of three neutral visual stimuli paired with a partial reinforcement by a mild electric current. We used both linear and non-linear dimensionality reduction approaches to distinguish between the initial RS and the RS after stimuli exposure. The principal component analysis (PCA) as a linear dimensionality reduction method showed significantly worse results than non-linear methods (Isomap, LLE, Laplacian eigenmaps). Using the Laplacian eigenmaps manifold learning method, we were able to show significant differences between the two RSs at the level of individual participants. This detection was further improved by smoothing the BOLD signal with the wavelet multiresolution analysis. The developed method can improve the discrimination of functional states collected in longitudinal fMRI studies.