Testing geometric representation hypotheses from simulated place cell recordings
This work addresses the challenge of interpreting neural encoding geometries in neuroscience, offering a method to better understand spatial representation in the brain, though it is incremental as it builds on existing manifold learning techniques.
The study tackled the problem of identifying whether hippocampal place cells encode spatial locations using Euclidean or graph-based geometries by simulating place cell recordings and applying manifold learning methods. The results showed that autoencoders accurately reflected the true geometric structure, outperforming PCA and UMAP, with autoencoders being more robust to noise.
Hippocampal place cells can encode spatial locations of an animal in physical or task-relevant spaces. We simulated place cell populations that encoded either Euclidean- or graph-based positions of a rat navigating to goal nodes in a maze with a graph topology, and used manifold learning methods such as UMAP and Autoencoders (AE) to analyze these neural population activities. The structure of the latent spaces learned by the AE reflects their true geometric structure, while PCA fails to do so and UMAP is less robust to noise. Our results support future applications of AE architectures to decipher the geometry of spatial encoding in the brain.