Artificial neural networks

We have entered a golden age of artificial intelligence research, driven mainly by the advances in the artificial neural networks over the last several decades. Applications of these techniques—to machine vision, speech recognition, autonomous vehicles, natural language, and many other domains—are coming so quickly that many observers predict that the long-elusive goal of “Artificial General Intelligence” (AGI) is within our grasp. However, we still cannot build a machine capable of building a nest, stalking prey, or loading a dishwasher. I will describe how evolution may have shaped the algor

Artificial neural networks are frequently used to model neural systems, and have been shown to recapitulate features of biological networks, including in the hippocampus [1] and visual cortex [2] [3]. However, despite it’s success in neural modelling, the predominant learning algorithm, backpropagation, has long been considered biologically implausible [4]. One reason is the weight-transport problem - or the problem of how a global error signal can be accurately transmitted across the network to minimize the error resulting from each layer’s parameters. Previous solutions to the weight-transpo

…information processing within neural circuits. Deficits in STP have been implicated in neurological and psychiatric disorders, but its precise contribution to higher cognitive functions and behavioral flexibility remains poorly understood. Here, we address this gap by focusing on synaptotagmin-7 (SYT7), a calcium sensor essential for facilitation---a form of STP in which high-frequency stimulation briefly enhances neurotransmitter release.

…This has inspired the development of artificial neural networks that learn representations by incorporating brain-like learning abilities. Recent work has focused on learning representations (latents) by capturing similarity between inputs and latents, but these methods have typically been used to learn downstream features from the input and have not been studied in the context of a generative paradigm, where one can map the representations back to the input space, incorporating not only bottom-

…Despite some progress in identifying neural correlates of specific social features, such as gaze and social rank, a comprehensive computational framework linking social perception to behavior remains elusive; at least in part due to the high dimensionality of both the input space (complex social scenes) and the output space (diverse behavioral

…has proven difficult to realize in artificial neural networks (ANNs) [1], which suffer from catastrophic forgetting of previously learned information. To understand this contrast, we sought to understand what aspects of a neural system and the tasks it learns are key to its CL performance. We developed a theory of deep, nonlinear feedforward networks sequentially learning multiple tasks and derived exact exp

Traditionally, the processing of object motion and form has been attributed to the dorsal and ventral visual pathways, respectively. However, recent studies challenge this strict dichotomy. For example, Hong et al. (2016) demonstrated that the ventral stream represents spatial information, while Ramezanpour et al. (2024) showed that object motion could be decoded from the inferior temporal (IT) cortex. These findings prompt further exploration of how the ventral pathway supports the integration of motion and form, particularly in naturalistic environments. In our study, we aimed to investigate

…systems neuroscience, and to interpret neural systems usi

The fact that the eyes are constantly in motion, even during ‘fixation’, implies that retinal ganglion cells' spike times carry information about the visual scene even when the scene is static. Moreover, this motion implies that fine details of the visual scene could not be decoded from pure spatial retinal representations due to smearing. Understanding the interplay of temporal and spatial information in visual processing is thus pivotal for both biological research and bio-inspired computer-vision applications. In this study, we consider data from an event-based camera that was designed to e

Artificial neural networks (ANNs) form the basis of most successful Deep Learning (DL) algorithms$^1$, which are capable of solving complex problems such as image recognition and natural language processing$^{2,3}$. However, unlike biological brains, which efficiently solve similar problems, DL algorithms require a large number of adjustable parameters, making them energy-intensive and susceptible to overfitting. In this study, a new ANN architecture is introduced that incorporates the structured connectivity and restricted sampling properties of biological dendrites, aiming to overcome these

…We used artificial neural networks to test the hypothesis that structural modularity is sufficient to guarantee functional specialization, and find that in general, this doesn't necessarily hold. We then systematically tested which features of the environment and network do lead to the emergence of specialization. We used a simple toy environment, task and network, allowing us precise control, and show that

Artificial neural networks (ANNs) achieve remarkable results on various tasks, but understanding the computational mechanisms underlying their performance remains difficult. Furthermore, traditionally employed artificial networks have little in common with real biological networks. Overcoming this difference, machine-learning-based training methods for spiking neuronal networks (SNNs) have been developed to create functional networks, enabling the investigation of these neural networks with neuroscientific analysis methods. Here we analyze one such SNN trained with backpropagation based on a

Recent progress in artificial intelligence (AI), particularly through the development of artificial neural networks (ANNs), has significantly benefited from insights gained from neuroscience. This progress has enhanced the replication of complex cognitive tasks such as vision and natural language processing (1,2). However, ANNs still struggle with continual learning, adaptable knowledge transfer, robustness, and resource efficiency - capabilities that biological systems handle seamlessly (3–8). Neuromodulators, the chemical messengers that target neurons and synapses in specific brain regions,

…In Artificial Neural Networks , this process faces the challenge of catastrophic forgetting, where updating parameters with new data causes an exponential decline in accuracy over time for previously learned tasks. Several methods have been proposed to address this problem and enable continual learning on future neuromorphic hardware. One such neuroscience-inspired method is metaplasticity, which refers to a network's ability to change the learning capacity of its synapses based on its cu

The success of Artificial Neural Networks (ANNs) trained with error-backpropagation algorithms relies on the combination of offline training, vast amounts of data, and energy-intensive simulations on dedicated hardware accelerators. In contrast, biological neurons can learn quickly in a self-supervised manner by utilizing (a) recurrent connections within their local network and from other areas, (b) segregated dendrites receiving inputs from feedforward and feedback pathways, and (c) learning rules that combine locally available information from pre- and -post-synaptic sources. Inspired by loc

Despite being inspired by biological neuronal networks (BNNs), artificial neural networks (ANNs) have fundamental differences from their biological counterparts. While ANNs excel at solving complex problems across various domains, BNNs outperform them in terms of energy efficiency and processing highly complex parallel inputs. However, the mechanisms behind these advantages remain unclear, making it challenging to integrate them into artificial computation. It has been shown that biological neurons exhibit a characteristic sigmoid-like firing rate response [1], termed bio-sigmoid, as a functi

We build upon and expand the efficient coding and predictive information models of neurons, presenting a novel perspective that neurons not only predict but also actively influence their future inputs through their outputs. We introduce the concept of neurons as feedback controllers of their environments, a role traditionally considered computationally demanding, particularly when the dynamical system characterizing the environment is unknown. By harnessing a novel data-driven control framework, we illustrate the feasibility of biological neurons functioning as effective feedback controllers.

Artificial neural networks (ANN) have emerged as computationally plausible models of human language processing. A major criticism of these models is that the amount of training data they receive far exceeds that of humans. Here, we use two complementary approaches to ask how the models’ ability to capture human neural and behavioral responses to language is affected by the amount of training data. First, we evaluate GPT-2 models trained on 1 million, 10 million, 100 million, or 1 billion tokens against a neural (fMRI) and a behavioral (reading times) benchmark. Because children are exposed to

…Models of visual cortex based on artificial neural networks (ANN) have been shown to contain internal representations that are remarkably similar to those observed in the visual cortex. However, unit selectivity in these models emerges without any particular spatial order (i.e. no topog

Working memory (WM) retains information during delay periods. How global neuronal activities maintain WM remains unclear. We recorded brain-wide activity via Neuropixels in mice performing a delay-varying olfactory WM task, to reveal animal’s capacity in maintaining both olfaction and duration information simultaneously. Neurons encoded WM information predominantly in transient patterns and generated cross-region activity waves, reflecting distinct cell assemblies that encoded olfaction-only, duration-only, or both information. The regional proportion of neurons and wave timing within cell ass

Predicting the future state of the environment is a crucial task for neural systems, which must compensate for significant delays in both sensation and action. Despite this fundamental importance, we lack a rigorous theoretical understanding of the underlying mechanisms of prediction. While several phenomenological and biophysical mechanisms have been proposed to model experimental data, they are typically nonlinear and difficult to analyze. On the other hand, prediction has been exhaustively studied in classical linear filtering theory, but this engineering approach does not explicitly addres

Cortical networks have the remarkable ability to self-assemble into dynamic regimes in which excitatory positive feedback is balanced by recurrent inhibition. This inhibition-stabilized regime is increasingly viewed as the default dynamic regime of the cortex but how it emerges in an unsupervised manner remains an open question. Theoretical work has established that four sets of weights (WE←E, WE←I, WI←E, WI←I) must obey specific relationships to produce inhibition-stabilized dynamics, but it is not known how the brain autonomously sets the values of all four weight classes to be in the suitab

…cortex (ITC), or its analogs in artificial neural networks. The leading hypothesis to account for this discrepancy proposes that the hippocampus (HC) applies a nonlinear transformation of its cortical inputs to better separate them before memory storage, a mechanism called hippocampal pattern separation (

…these priors can be set nor how neural activity underlies these fundamental abilities. To answer this, we designed a reaching task in which subjects (human or m