Identification of recurrent dynamics in distributed neural populations

PLoS Computational Biology, Journal Year: 2025, Volume and Issue: 21(2), P. e1012816 - e1012816

Published: Feb. 6, 2025

Large-scale recordings of neural activity over broad anatomical areas with high spatial and temporal resolution are increasingly common in modern experimental neuroscience. Recently, recurrent switching dynamical systems have been used to tackle the scale complexity these data. However, an important challenge remains providing insights into existence structure linear dynamics time series Here we test a scalable approach time-varying autoregression low-rank tensors recover stochastic mass models multiple stable attractors. We demonstrate that parsimonious representation system matrices terms modes can attractor simple via clustering. then consider simulations based on human brain connectivity matrix low global connection strength regimes, reveal hierarchical clustering dynamics. Finally, explain impact forecast delay estimation underlying rank variability This study illustrates prediction error minimization is not sufficient meaningful dynamic it crucial account for three key timescales arising from dynamics, noise processes, switching.

Language: Английский

---

Computation-through-Dynamics Benchmark: Simulated datasets and quality metrics for dynamical models of neural activity

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 8, 2025

A primary goal of systems neuroscience is to discover how ensembles neurons transform inputs into goal-directed behavior, a process known as neural computation. powerful framework for understanding computation uses dynamics - the rules that describe temporal evolution activity explain input-output transformations occur. As dynamical are not directly observable, we need computational models can infer from recorded activity. We typically validate such using synthetic datasets with ground-truth dynamics, but unfortunately existing don reflect fundamental features and thus poor proxies systems. Further, field lacks validated metrics quantifying accuracy inferred by models. The Computation-through-Dynamics Benchmark (CtDB) fills these critical gaps providing: 1) properties biological circuits, 2) interpretable model performance, 3) standardized pipeline training evaluating or without external inputs. In this manuscript, demonstrate CtDB help guide development, tuning, troubleshooting summary, provides platform developers better understand characterize through lens dynamics.

Language: Английский

---

Focal Infrared Neural Stimulation Propagates Dynamical Transformations in Auditory Cortex

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2025, Volume and Issue: unknown

Published: March 13, 2025

Abstract Significance Infrared neural stimulation (INS) has emerged as a potent neuromodulation technology, offering safe and focal with superior spatial recruitment profiles compared to conventional electrical methods. However, the dynamics induced by INS remain poorly understood. Elucidating these will help develop new paradigms advance its clinical application. Aim In this study, we assessed local network of entrainment in auditory thalamocortical circuit using chronically implanted rat model; our approach focused on measuring energy-based field potential (LFP) stimulation. We further characterized linear nonlinear oscillatory LFP activity response single-pulse periodic performed spectral decomposition uncover specific band INS. Finally, examined spike-field transformations across synapse spike-LFP coherence coupling. Results found that significantly increases amplitude log-linear function energy per pulse, primarily entraining β γ bands synchrony extending 200 Hz some cases. A subset neurons demonstrated nonlinear, chaotic oscillations linked information transfer cortical circuits. utilized coherences correlate spike coupling frequency suggest an energy-dependent model activation resulting from Conclusions show reliably drives robust can potently modulate potentials wide range frequencies stimulus parameter-dependent manner. Based results, propose design principles for developing full coverage, all-optical neuroprostheses.

Language: Английский

---

The butterfly effect in neural networks: Unveiling hyperbolic chaos through parameter sensitivity

Neural Networks, Journal Year: 2025, Volume and Issue: 189, P. 107572 - 107572

Published: May 15, 2025

Language: Английский

---

Protected Chaos in a Topological Lattice

Advanced Science, Journal Year: 2025, Volume and Issue: unknown

Published: May 20, 2025

Abstract The erratic nature of chaotic behavior is thought to erode the stability periodic behavior, including topological oscillations. However, it discovered that in presence chaos, non‐trivial topology not only endures but also provides robust protection dynamics within a lattice hosting non‐linear oscillators. Despite difficulty defining invariants settings, robustness still persists parametric state boundary This interplay between chaos and demonstrated by incorporating Chua's circuits into Su‐Schrieffer‐Heeger (SSH) circuit. By extrapolating from linear limit deep regime, found distinctive correlations bulk edge scroll effectively capture origin protected chaos. findings suggest topologically can be robustly achieved across broad spectrum periodically driven systems, thereby offering new avenues for design resilient adaptable networks.

Language: Английский

---

The construction method of chaotic system model based on state variables and uncertain variables and its application in image encryption

Applied Mathematical Modelling, Journal Year: 2025, Volume and Issue: unknown, P. 116097 - 116097

Published: March 1, 2025

Language: Английский

---

Multi-task and continual-learning capabilities of reservoir computing applied to speech recognition

2022 International Joint Conference on Neural Networks (IJCNN), Journal Year: 2024, Volume and Issue: unknown, P. 1 - 6

Published: June 30, 2024

Language: Английский

---

Chaotic recurrent neural networks for brain modelling: A review

Neural Networks, Journal Year: 2024, Volume and Issue: 184, P. 107079 - 107079

Published: Dec. 27, 2024

Language: Английский

---