This paper provides a perspective on applying the concepts of information thermodynamics, developed recently in non-equilibrium statistical physics, to problems in theoretical neuroscience. Historically, information and energy in neuroscience have been treated separately, in contrast to physics approaches, where the relationship of entropy production with heat is a central idea. It is argued here that also in neural systems information and energy can be considered within the same theoretical framework. Starting from basic ideas of thermodynamics and information theory on a classic Brownian particle, it is shown how noisy neural networks can infer its probabilistic motion. The decoding of the particle motion by neurons is performed with some accuracy and it has some energy cost, and both can be determined using information thermodynamics. In a similar fashion, we also discuss how neural networks in the brain can learn the particle velocity, and maintain that information in the weights of plastic synapses from a physical point of view. Generally, it is shown how the framework of stochastic and information thermodynamics can be used practically to study neural inference, learning, and
The brain can learn to execute a wide variety of tasks quickly and efficiently. Nevertheless, most of the mechanisms that enable us to learn are unclear or incredibly complicated. Recently, considerable efforts have been made in neuroscience and artificial intelligence to understand and model the structure and mechanisms behind the amazing learning capability of the brain. However, in the current understanding of cognitive neuroscience, it is widely accepted that synaptic plasticity plays an essential role in our amazing learning capability. This mechanism is also known as the Credit Assignment Problem (CAP) and is a fundamental challenge in neuroscience and Artificial Intelligence (AI). The observations of neuroscientists clearly confirm the role of two important mechanisms including the error feedback system and unsupervised learning in synaptic plasticity. With this inspiration, a new learning rule is proposed via the fusion of reinforcement learning (RL) and unsupervised learning (UL). In the proposed computational model, the nonlinear optimal control theory is used to resemble the error feedback loop systems and project the output error to neurons membrane potential (neurons s
The endeavor to understand the brain involves multiple collaborating research fields. Classically, synaptic plasticity rules derived by theoretical neuroscientists are evaluated in isolation on pattern classification tasks. This contrasts with the biological brain which purpose is to control a body in closed-loop. This paper contributes to bringing the fields of computational neuroscience and robotics closer together by integrating open-source software components from these two fields. The resulting framework allows to evaluate the validity of biologically-plausibe plasticity models in closed-loop robotics environments. We demonstrate this framework to evaluate Synaptic Plasticity with Online REinforcement learning (SPORE), a reward-learning rule based on synaptic sampling, on two visuomotor tasks: reaching and lane following. We show that SPORE is capable of learning to perform policies within the course of simulated hours for both tasks. Provisional parameter explorations indicate that the learning rate and the temperature driving the stochastic processes that govern synaptic learning dynamics need to be regulated for performance improvements to be retained. We conclude by discus
Neural plasticity is fundamental to memory storage and retrieval in biological systems, yet existing models often fall short in addressing noise sensitivity and unbounded synaptic weight growth. This paper investigates the Allee-based nonlinear plasticity model, emphasizing its biologically inspired weight stabilization mechanisms, enhanced noise robustness, and critical thresholds for synaptic regulation. We analyze its performance in memory retention and pattern retrieval, demonstrating increased capacity and reliability compared to classical models like Hebbian and Oja's rules. To address temporal limitations, we extend the model by integrating time-dependent dynamics, including eligibility traces and oscillatory inputs, resulting in improved retrieval accuracy and resilience in dynamic environments. This work bridges theoretical insights with practical implications, offering a robust framework for modeling neural adaptation and informing advances in artificial intelligence and neuroscience.
The brain is a nonlinear and highly Recurrent Neural Network (RNN). This RNN is surprisingly plastic and supports our astonishing ability to learn and execute complex tasks. However, learning is incredibly complicated due to the brain's nonlinear nature and the obscurity of mechanisms for determining the contribution of each synapse to the output error. This issue is known as the Credit Assignment Problem (CAP) and is a fundamental challenge in neuroscience and Artificial Intelligence (AI). Nevertheless, in the current understanding of cognitive neuroscience, it is widely accepted that a feedback loop systems play an essential role in synaptic plasticity. With this as inspiration, we propose a computational model by combining Neural Networks (NN) and nonlinear optimal control theory. The proposed framework involves a new NN-based actor-critic method which is used to simulate the error feedback loop systems and projections on the NN's synaptic plasticity so as to ensure that the output error is minimized.
Like many scientific disciplines, neuroscience has increasingly attempted to confront pervasive gender imbalances within the field. While much of the conversation has centered around publishing and conference participation, recent research in other fields has called attention to the prevalence of gender bias in citation practices. Because of the downstream effects that citations can have on visibility and career advancement, understanding and eliminating gender bias in citation practices is vital for addressing inequity in a scientific community. In this study, we sought to determine whether there is evidence of gender bias in the citation practices of neuroscientists. Using data from five top neuroscience journals, we find that reference lists tend to include more papers with men as first and last author than would be expected if gender were not a factor in referencing. Importantly, we show that this overcitation of men and undercitation of women is driven largely by the citation practices of men, and is increasing over time as the field becomes more diverse. We develop a co-authorship network to assess homophily in researchers' social networks, and we find that men tend to overci
Learning, especially rapid learning, is critical for survival. However, learning is hard: a large number of synaptic weights must be set based on noisy, often ambiguous, sensory information. In such a high-noise regime, keeping track of probability distributions over weights is the optimal strategy. Here we hypothesize that synapses take that strategy; in essence, when they estimate weights, they include error bars. They then use that uncertainty to adjust their learning rates, with more uncertain weights having higher learning rates. We also make a second, independent, hypothesis: synapses communicate their uncertainty by linking it to variability in PSP size, with more uncertainty leading to more variability. These two hypotheses cast synaptic plasticity as a problem of Bayesian inference, and thus provide a normative view of learning. They generalize known learning rules, offer an explanation for the large variability in the size of post-synaptic potentials, and make falsifiable experimental predictions.
Spiking Neural Networks (SNNs), as one of the algorithmic models in neuromorphic computing, have gained a great deal of research attention owing to temporal information processing capability, low power consumption, and high biological plausibility. The potential to efficiently extract spatio-temporal features makes it suitable for processing the event streams. However, existing synaptic structures in SNNs are almost full-connections or spatial 2D convolution, neither of which can extract temporal dependencies adequately. In this work, we take inspiration from biological synapses and propose a spatio-temporal synaptic connection SNN (STSC-SNN) model, to enhance the spatio-temporal receptive fields of synaptic connections, thereby establishing temporal dependencies across layers. Concretely, we incorporate temporal convolution and attention mechanisms to implement synaptic filtering and gating functions. We show that endowing synaptic models with temporal dependencies can improve the performance of SNNs on classification tasks. In addition, we investigate the impact of performance vias varied spatial-temporal receptive fields and reevaluate the temporal modules in SNNs. Our approach
Recent advances and reflections on reproducible human neuroscience, especially brain-wide association studies (BWAS) leveraging large datasets, have led to divergent and sometimes opposing views on research practices and priorities. The debates span multiple dimensions. Shifts along these axes have fractured consensus and further fragmented an already heterogeneous field of cognitive neuroscience. Here, we sketch a holistic and integrative response grounded in population neuroscience, organized around a closed-loop "design-analysis-interpretation" research cycle that aims to build consensus while bridging these divides. Our central claim is that population neuroscience offers a unique population-level vantage point for identifying general principles, characterizing inter-individual variabilities, and benchmarking intra-individual changes, thereby providing a supportive framework for small-scale, mechanism-focused studies at the individual level and allowing them to co-evolve with population-level studies. Population neuroscience is not simply about providing larger N for BWAS; its deeper goal is to accumulate a family of cross-scale priors and shared infrastructures that can suppor
Neuromorphic Computing (NC), which emulates neural activities of the human brain, is considered for low-power implementation of artificial intelligence. Towards realizing NC, fabrication, and investigations of hardware elements such as synaptic devices and neurons are essential. Electrolyte gating has been widely used for conductance modulation by massive carrier injections and has proven to be an effective way of emulating biological synapses. Synaptic devices, in the form of synaptic transistors, have been studied using a wide variety of materials. However, studies on metallic channel based synaptic transistors remain vastly unexplored. Here, we have demonstrated a three-terminal cobalt-based synaptic transistor to emulate biological synapse. We realized gating controlled multilevel, nonvolatile conducting states in the proposed device. The device could successfully emulate essential synaptic functions demonstrating short-term and long-term plasticity. A transition from short-term memory to long-term memory has been realized by tuning gate pulse amplitude and duration. The crucial cognitive behavior viz., learning, forgetting, and relearning, has been emulated, showing resemblanc
The proliferation and refinement of affordable virtual reality (VR) technologies and wearable sensors have opened new frontiers in cognitive and behavioral neuroscience. This chapter offers a broad overview of VR for anyone interested in leveraging it as a research tool. In the first section, it examines the fundamental functionalities of VR and outlines important considerations that inform the development of immersive content that stimulates the senses. In the second section, the focus of the discussion shifts to the implementation of VR in the context of the neuroscience lab. Practical advice is offered on adapting commercial, off-theshelf devices to specific research purposes. Further, methods are explored for recording, synchronizing, and fusing heterogeneous forms of data obtained through the VR system or add-on sensors, as well as for labeling events and capturing game play.
A fundamental question in neuroscience is how structure and function of neural systems are related. We study this interplay by combining a familiar auto-associative neural network with an evolving mechanism for the birth and death of synapses. A feedback loop then arises leading to two qualitatively different types of behaviour. In one, the network structure becomes heterogeneous and dissasortative, and the system displays good memory performance; furthermore, the structure is optimised for the particular memory patterns stored during the process. In the other, the structure remains homogeneous and incapable of pattern retrieval. These findings provide an inspiring picture of brain structure and dynamics, are compatible with experimental results on early brain development, and may help to explain synaptic pruning. Other evolving networks -- such as those of protein interaction -- might share the basic ingredients for this feedback loop and other questions, and indeed many of their structural features are as predicted by our model.
A growing body of work underlines striking similarities between biological neural networks and recurrent, binary neural networks. A relatively smaller body of work, however, discusses similarities between learning dynamics employed in deep artificial neural networks and synaptic plasticity in spiking neural networks. The challenge preventing this is largely caused by the discrepancy between the dynamical properties of synaptic plasticity and the requirements for gradient backpropagation. Learning algorithms that approximate gradient backpropagation using locally synthesized gradients can overcome this challenge. Here, we show that synthetic gradients enable the derivation of Deep Continuous Local Learning (DECOLLE) in spiking neural networks. DECOLLE is capable of learning deep spatio-temporal representations from spikes relying solely on local information. Synaptic plasticity rules are derived systematically from user-defined cost functions and neural dynamics by leveraging existing autodifferentiation methods of machine learning frameworks. We benchmark our approach on the MNIST and the event-based neuromorphic DvsGesture dataset, on which DECOLLE performs comparably to the state
Benchmarking the performance of complex systems such as rail networks, renewable generation assets and national economies is central to transport planning, regulation and macroeconomic analysis. Classical frontier methods, notably Data Envelopment Analysis (DEA) and Stochastic Frontier Analysis (SFA), estimate an efficient frontier in the observed input-output space and define efficiency as distance to this frontier, but rely on restrictive assumptions on the production set and only indirectly address heterogeneity and scale effects. We propose Geometric Manifold Analysis (GeMA), a latent manifold frontier framework implemented via a productivity-manifold variational autoencoder (ProMan-VAE). Instead of specifying a frontier function in the observed space, GeMA represents the production set as the boundary of a low-dimensional manifold embedded in the joint input-output space. A split-head encoder learns latent variables that capture technological structure and operational inefficiency. Efficiency is evaluated with respect to the learned manifold, endogenous peer groups arise as clusters in latent technology space, a quotient construction supports scale-invariant benchmarking, and
The representation of the natural-density, heterogeneous connectivity of neuronal network models at relevant spatial scales remains a challenge for Computational Neuroscience and Neuromorphic Computing. In particular, the memory demands imposed by the vast number of synapses in brain-scale network simulations constitutes a major obstacle. Limiting the number resolution of synaptic weights appears to be a natural strategy to reduce memory and compute load. In this study, we investigate the effects of a limited synaptic-weight resolution on the dynamics of recurrent spiking neuronal networks resembling local cortical circuits, and develop strategies for minimizing deviations from the dynamics of networks with high-resolution synaptic weights. We mimic the effect of a limited synaptic weight resolution by replacing normally distributed synaptic weights by weights drawn from a discrete distribution, and compare the resulting statistics characterizing firing rates, spike-train irregularity, and correlation coefficients with the reference solution. We show that a naive discretization of synaptic weights generally leads to a distortion of the spike-train statistics. Only if the weights ar
Many mathematical models of synaptic plasticity have been proposed to explain the diversity of plasticity phenomena observed in biological organisms. These models range from simple interpretations of Hebb's postulate, which suggests that correlated neural activity leads to increases in synaptic strength, to more complex rules that allow bidirectional synaptic updates, ensure stability, or incorporate additional signals like reward or error. At the same time, a range of learning paradigms can be observed behaviorally, from Pavlovian conditioning to motor learning and memory recall. Although it is difficult to directly link synaptic updates to learning outcomes experimentally, computational models provide a valuable tool for building evidence of this connection. In this chapter, we discuss several fundamental learning paradigms, along with the synaptic plasticity rules that might be used to implement them.
Synaptic plasticity is vital for learning and memory in the brain. It consists of long-term potentiation (LTP) and long-term depression (LTD). Spike frequency is one of the major components of synaptic plasticity in the brain, a noisy environment. Recently, we mathematically analysed the frequency-dependent synaptic plasticity (FDP) in vivo and found that LTP is more likely to occur with an increase in the frequency of background synaptic activity. Previous studies suggest fluctuation in the amplitude of background synaptic activity. However, little is understood about the relationship between synaptic plasticity and the fluctuation in the background synaptic activity. To address this issue, we performed numerical simulations of a calcium-based synapse model. Then, we found attenuation of the tendency to become LTD due to an increase in the fluctuation of background synaptic activity, leading to an enhancement of synaptic weight. Our result suggests that the fluctuation affect synaptic plasticity in the brain.
While probabilistic models describe the dependence structure between observed variables, causal models go one step further: they predict, for example, how cognitive functions are affected by external interventions that perturb neuronal activity. In this review and perspective article, we introduce the concept of causality in the context of cognitive neuroscience and review existing methods for inferring causal relationships from data. Causal inference is an ambitious task that is particularly challenging in cognitive neuroscience. We discuss two difficulties in more detail: the scarcity of interventional data and the challenge of finding the right variables. We argue for distributional robustness as a guiding principle to tackle these problems. Robustness (or invariance) is a fundamental principle underlying causal methodology. A causal model of a target variable generalises across environments or subjects as long as these environments leave the causal mechanisms intact. Consequently, if a candidate model does not generalise, then either it does not consist of the target variable's causes or the underlying variables do not represent the correct granularity of the problem. In this s
Short-term memory in the brain cannot in general be explained the way long-term memory can -- as a gradual modification of synaptic weights -- since it takes place too quickly. Theories based on some form of cellular bistability, however, do not seem able to account for the fact that noisy neurons can collectively store information in a robust manner. We show how a sufficiently clustered network of simple model neurons can be instantly induced into metastable states capable of retaining information for a short time (a few seconds). The mechanism is robust to different network topologies and kinds of neural model. This could constitute a viable means available to the brain for sensory and/or short-term memory with no need of synaptic learning. Relevant phenomena described by neurobiology and psychology, such as local synchronization of synaptic inputs and power-law statistics of forgetting avalanches, emerge naturally from this mechanism, and we suggest possible experiments to test its viability in more biological settings.
Synaptic efficacy between neurons is known to change within a short time scale dynamically. Neurophysiological experiments show that high-frequency presynaptic inputs decrease synaptic efficacy between neurons. This phenomenon is called synaptic depression, a short term synaptic plasticity. Many researchers have investigated how the synaptic depression affects the memory storage capacity. However, the noise has not been taken into consideration in their analysis. By introducing "temperature", which controls the level of the noise, into an update rule of neurons, we investigate the effects of synaptic depression on the memory storage capacity in the presence of the noise. We analytically compute the storage capacity by using a statistical mechanics technique called Self Consistent Signal to Noise Analysis (SCSNA). We find that the synaptic depression decreases the storage capacity in the case of finite temperature in contrast to the case of the low temperature limit, where the storage capacity does not change.