George, Tom M.;
(2025)
Computational Models of Learning and Representations in the Hippocampal Formation.
Doctoral thesis (Ph.D), UCL (University College London).
Preview |
Text
George_10214381_thesis_sigs_removed.pdf Download (26MB) | Preview |
Abstract
This thesis explores how neural systems learn and use internal representations to support flexible behaviour, focusing on the mammalian spatial memory system and cognitive map. It introduces new computational tools and biologically plausible models that link neural structure and dynamics to spatial cognition. Foundational to this work is RatInABox, an open-source toolkit now widely used for simulating realistic navigation and hippocampal activity. This platform enables the rapid prototyping of models that jointly capture behavioural trajectories and neural representations, including place and grid cells. The first model leverages this toolkit to demonstrate a biologically plausible mechanism for learning predictive representations via spike-timing dependent plasticity and theta phase precession. This mechanism bridges the timescale gap between behaviour and synaptic plasticity to enable fast and flexible learning. A second study introduces a generative model of the hippocampal-entorhinal loop that unifies path integration and mental simulation. It shows how local Hebbian learning, scheduled by theta oscillations, can give rise to ring attractor dynamics within a normative Helmholtz machine framework. The final contribution is SIMPL, an efficient algorithm for latent variable discovery from high-dimensional neural data. By recursively fitting latent trajectories and tuning curves, SIMPL achieves state-of-the-art performance while enhancing the interpretability and precision of neural representations. Collectively, these contributions advance our understanding of how biological systems learn and represent the world, and provide new models and tools for research at the intersection of neuroscience and artificial intelligence.
| Type: | Thesis (Doctoral) |
|---|---|
| Qualification: | Ph.D |
| Title: | Computational Models of Learning and Representations in the Hippocampal Formation |
| Open access status: | An open access version is available from UCL Discovery |
| Language: | English |
| Additional information: | Copyright © The Author 2025. Original content in this thesis is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) Licence (https://creativecommons.org/licenses/by-nc/4.0/). Any third-party copyright material present remains the property of its respective owner(s) and is licensed under its existing terms. Access may initially be restricted at the author’s request. |
| UCL classification: | UCL UCL > Provost and Vice Provost Offices > School of Life and Medical Sciences UCL > Provost and Vice Provost Offices > School of Life and Medical Sciences > Faculty of Life Sciences UCL > Provost and Vice Provost Offices > School of Life and Medical Sciences > Faculty of Life Sciences > The Sainsbury Wellcome Centre |
| URI: | https://discovery.ucl.ac.uk/id/eprint/10214381 |
Archive Staff Only
 |
View Item |
