Balancing stability and flow in hippocampal networks via inductive bias and learned symmetry breaking
Philosophical Transactions B · 09.07.2026 · 381 (1954)
DOI: 10.1098/rstb.2025.0260 ↗Kurz gefasst
Fachartikel in der Zeitschrift Philosophical Transactions B, online erschienen am 09.07.2026. Verfasst von Margot Wagner und 4 weiteren Autor:innen. Beteiligte Einrichtung: Institute for Neural Computation, University of California San Diego, La Jolla, CA, USA. Hauptthemen laut MEDLINE: Hippocampus, Learning, Memory, Nerve Net. Die Arbeit nennt 102 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: U.S. Naval Research Laboratory, NIH Office of the Director.
Abstract (Originalsprache)
The hippocampus must balance stable memory representations with internally generated sequential dynamics underlying replay and prediction. How hippocampal circuitry achieves both remains unclear. Here, we show that recurrent neural networks trained on prediction tasks converge to a mixed-symmetry dynamical regime, in which dominant symmetric recurrence stabilizes an attractor while a weaker antisymmetric component induces directed flow. This structure accelerates learning and supports robust replay and prediction. We further show that such symmetry breaking can arise from biologically plausible spike-timing-dependent plasticity rules, yielding a tilted Mexican-hat connectivity profile without explicit architectural constraints. Initializing networks with CA3-like structured connectivity biases learning towards this regime and improves optimization efficiency and performance. These results suggest that hippocampal computation reflects an interaction between biologically constrained circuit structure and task-driven learning, with partial symmetry breaking providing a low-dimensional control principle for balancing stability and flow in sequence generation.This article is part of the theme issue ‘The role of hippocampal predictions in cognition: bridging perception and memory’.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Institute for Neural Computation, University of California San Diego, La Jolla, CA, USA.
- Institute of Neuroscience, Chinese Academy of Sciences, Beijing, Beijing, People's Republic of China.
- Computational Neurobiology Laboratory, Salk Institute for Biological Studies, La Jolla, CA, USA.
- Computational Neuroscience Center, University of Washington, Seattle, WA, USA.
ORCID-Profile: Margot Wagner
Förderung
- U.S. Naval Research Laboratory
- NIH Office of the Director
Zitieren
APA
Wagner, M., Chen, Y., Karuvally, A., Cameron, M., & Sejnowski, T. (2026). Balancing stability and flow in hippocampal networks via inductive bias and learned symmetry breaking. Philosophical Transactions B, 381(1954). https://doi.org/10.1098/rstb.2025.0260
Vancouver
Wagner M, Chen Y, Karuvally A, Cameron M, Sejnowski T. Balancing stability and flow in hippocampal networks via inductive bias and learned symmetry breaking. Philosophical Transactions B. 2026;381(1954). doi:10.1098/rstb.2025.0260
BibTeX
@article{Wagner2026,
author = {Wagner, Margot and Chen, Yusi and Karuvally, Arjun and Cameron, Mia and Sejnowski, Terry},
title = {Balancing stability and flow in hippocampal networks via inductive bias and learned symmetry breaking},
journal = {Philosophical Transactions B},
year = {2026},
volume = {381},
number = {1954},
doi = {10.1098/rstb.2025.0260}
}
EUVIMED