EUVIMEDEUVIMED entdecken
Crossref

Stiefel Manifold Dynamical Systems for Tracking Representational Drift

Hyun Dong Lee, Aditi Jha, Stephen E. Clarke, Michael P. Silvernagel, Paul Nuyujukian, Scott W. Linderman

Neural Computation · 05.11.2026 · 1-19

DOI: 10.1162/neco.a.1599 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Neural Computation, online erschienen am 05.11.2026. Verfasst von Hyun Dong Lee und 5 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Understanding neural dynamics is crucial for uncovering how the brain processes information and controls behavior. Linear dynamical systems (LDS) are widely used for modeling neural data due to their simplicity and effectiveness in capturing latent dynamics. However, LDS assume a stable mapping from the latent states to neural activity, limiting its ability to capture representational drift—gradual changes in the brain’s representation of the external world. To address this, we introduce the Stiefel manifold dynamical system (SMDS), a new class of model designed to account for drift in neural representations across trials. In SMDS, emission matrices are constrained to be orthonormal and evolve smoothly over trials on the Stiefel manifold—the space of all orthonormal matrices—while the dynamics parameters are shared. This formulation allows SMDS to leverage data across trials while accounting for nonstationarity, thus capturing the underlying neural dynamics more accurately compared to an LDS. We apply SMDS to both simulated data sets and neural recordings across species. Our results consistently show that SMDS outperforms LDS in terms of log-likelihood and requires fewer latent dimensions to capture the same activity. Moreover, SMDS provides a powerful framework for quantifying and interpreting representational drift. It reveals a gradual drift over the course of minutes in the neural recordings and uncovers varying drift rates across dimensions, with slower drift in behaviorally and neurally significant dimensions.

Zitieren

APA

Lee, H. D., Jha, A., Clarke, S. E., Silvernagel, M. P., Nuyujukian, P., & Linderman, S. W. (2026). Stiefel Manifold Dynamical Systems for Tracking Representational Drift. Neural Computation, 1-19. https://doi.org/10.1162/neco.a.1599

Vancouver

Lee HD, Jha A, Clarke SE, Silvernagel MP, Nuyujukian P, Linderman SW. Stiefel Manifold Dynamical Systems for Tracking Representational Drift. Neural Computation. 2026:1-19. doi:10.1162/neco.a.1599

BibTeX
@article{Lee2026,
  author = {Lee, Hyun Dong and Jha, Aditi and Clarke, Stephen E. and Silvernagel, Michael P. and Nuyujukian, Paul and Linderman, Scott W.},
  title = {Stiefel Manifold Dynamical Systems for Tracking Representational Drift},
  journal = {Neural Computation},
  year = {2026},
  pages = {1--19},
  doi = {10.1162/neco.a.1599}
}
Kostenlos starten →