Geometry of abstract learned knowledge in the hippocampus

• Published in: Nature (London) Vol. 595; no. 7865; pp. 80 - 84
• Main Authors: Nieh, Edward H., Schottdorf, Manuel, Freeman, Nicolas W., Low, Ryan J., Lewallen, Sam, Koay, Sue Ann, Pinto, Lucas, Gauthier, Jeffrey L., Brody, Carlos D., Tank, David W.
• Format: Journal Article
• Language: English
• Published: London Nature Publishing Group UK 01.07.2021; Nature Publishing Group
• Subjects: 14/69; 631/378; 631/378/3920; 64/60; Animals; CA1 Region, Hippocampal - cytology; CA1 Region, Hippocampal - physiology; Calcium - metabolism; Calcium imaging; Cognitive ability; Cognitive maps; Cognitive models; Decision Making; Female; Functional magnetic resonance imaging; Hippocampus; Hippocampus - cytology; Hippocampus - physiology; Humanities and Social Sciences; Knowledge; Knowledge representation; Learning - physiology; Magnetic resonance; Magnetic resonance imaging; Male; Manifolds (mathematics); Mental task performance; Mice; Models, Neurological; multidisciplinary; Neurons; Neurons - metabolism; Population; Principal components analysis; Science; Science (multidisciplinary); Variables
• ISSN: 0028-0836; 1476-4687
• DOI: 10.1038/s41586-021-03652-7

Hippocampal neurons encode physical variables 1 – 7 such as space 1 or auditory frequency 6 in cognitive maps 8 . In addition, functional magnetic resonance imaging studies in humans have shown that the hippocampus can also encode more abstract, learned variables 9 – 11 . However, their integration into existing neural representations of physical variables 12 , 13 is unknown. Here, using two-photon calcium imaging, we show that individual neurons in the dorsal hippocampus jointly encode accumulated evidence with spatial position in mice performing a decision-making task in virtual reality 14 – 16 . Nonlinear dimensionality reduction 13 showed that population activity was well-described by approximately four to six latent variables, which suggests that neural activity is constrained to a low-dimensional manifold. Within this low-dimensional space, both physical and abstract variables were jointly mapped in an orderly manner, creating a geometric representation that we show is similar across mice. The existence of conjoined cognitive maps suggests that the hippocampus performs a general computation—the creation of task-specific low-dimensional manifolds that contain a geometric representation of learned knowledge. The hippocampus geometrically represents both physical location and abstract variables on a neural manifold in mice performing a decision-making task in virtual reality.