Inferring Multi-Timescale Neural Dynamics with Switching Linear Dynamical Systems
Abstract
Neural activity often exhibits multiple timescales that can vary with behavioral states and task conditions. Identifying these timescales from neural recordings is important for better understanding neural computation and function. However, traditional approaches based on autocorrelation fitting are difficult to scale to high-dimensional population recordings and can become unreliable when neural dynamics change with behavior. State-space models have been a powerful framework for modeling high-dimensional neural population activity through latent dynamical systems, but standard formulations and inference methods do not explicitly account for multiple timescales and therefore do not guarantee accurate recovery of the underlying temporal structure. Motivated by these questions, we introduce the Multi-Timescale Switching Linear Dynamical System (MTS-SLDS), a framework for identifying regime-specific latent timescales from continuous or spiking neural observations. MTS-SLDS combines a multi-lag moment initialization, which captures temporal structure across multiple observation lags, with \textit{regime-conditioned} Laplace-EM inference, which reduces mixing of dynamical statistics across uncertain regimes. Characteristic timescales can then be extracted directly from the eigenvalues of the learned latent transition matrices. In synthetic and neural experiments with Gaussian and Poisson spike observations, MTS-SLDS accurately recovers timescales and switching structure over multiple datasets.
Related
- Inferring structural complexity from dynamical measures
- Plasticity Dynamics Improves Reservoir Performance on Mixed-Timescale Memory Tasks
- Long-timescale dynamics shape visual conscious awareness
- Recording proteome dynamics in the late-stage embryo at hour-timescale, spatial, and single-cell resolution with MEMBRYO
- B-GRASP: A Bayesian Framework for Inferring Graph Weights from SPDE-Inspired Dynamics
BibTeX
@misc{gong2026inferring,
title = {{Inferring Multi-Timescale Neural Dynamics with Switching Linear Dynamical Systems}},
author = {Lulu Gong and Yongxu Zhang and Shreya Saxena},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.01786},
eprint = {2610.01786},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.01786},
}
RIS
TY - UNPB
TI - Inferring Multi-Timescale Neural Dynamics with Switching Linear Dynamical Systems
AU - Lulu Gong
AU - Yongxu Zhang
AU - Shreya Saxena
PY - 2026
DA - 2026/10/01
PB - arXiv
DO - 10.48550/arxiv.2610.01786
UR - https://arxiv.org/abs/2610.01786
AB - Neural activity often exhibits multiple timescales that can vary with behavioral states and task conditions. Identifying these timescales from neural recordings is important for better understanding neural computation and function. However, traditional approaches based on autocorrelation fitting are difficult to scale to high-dimensional population recordings and can become unreliable when neural dynamics change with behavior. State-space models have been a powerful framework for modeling high-dimensional neural population activity through latent dynamical systems, but standard formulations and inference methods do not explicitly account for multiple timescales and therefore do not guarantee accurate recovery of the underlying temporal structure. Motivated by these questions, we introduce the Multi-Timescale Switching Linear Dynamical System (MTS-SLDS), a framework for identifying regime-specific latent timescales from continuous or spiking neural observations. MTS-SLDS combines a multi-lag moment initialization, which captures temporal structure across multiple observation lags, with \textit{regime-conditioned} Laplace-EM inference, which reduces mixing of dynamical statistics across uncertain regimes. Characteristic timescales can then be extracted directly from the eigenvalues of the learned latent transition matrices. In synthetic and neural experiments with Gaussian and Poisson spike observations, MTS-SLDS accurately recovers timescales and switching structure over multiple datasets.
ER -