Caswell Barry

The paper investigates the neuronal origin of hexadirectional modulation in grid cells using data from foraging rats, finding that firing rate adaptation explains the phenomenon.

The paper shows that left-right alternation in hippocampal theta sequences is driven by medial entorhinal cortex dynamics, not intentional planning.

This study shows that subicular neurons encode predictive spatial representations by integrating behavioral biases into the successor representation framework, explaining their diverse activity patterns.

This paper uses recurrent neural networks to show how changes in locomotor movement statistics during development shape the emergence of spatial tuning in hippocampal neurons.

This paper investigates how symmetric learning rules affect predictive neural representations and generalization in spatial navigation using a neural network model of hippocampal subfields.

This study shows that disrupted hippocampal reactivations in AD mice predict spatial memory deficits by reducing place cell stability.

SIMPL optimizes neural representations of behaviour by iteratively refining latent variables and tuning curves, revealing more precise spatial encoding than traditional methods.

Repix enables reliable, reusable, and versatile chronic Neuropixels recordings in unrestrained animals over months, with minimal user training and high stability.

NeuralPlayground provides a standardized framework to compare computational models of hippocampus and entorhinal cortex with experimental data on spatial cognition.

RatInABox generates synthetic locomotion and neural data for studying spatial navigation models in rodents.

This study investigates how theta-band phase locking between entorhinal and hippocampal neurons supports memory consolidation during replay.

The paper shows that theta phase precession enhances memorability of sensory-driven vector fields and persists in memory-dependent fields during retrieval.

This paper investigates how environment geometry affects subiculum boundary vector cells in adults and early development, revealing geometry-dependent tuning and inhibitory contributions to receptive fields.

This paper demonstrates that spike-timing dependent plasticity (STDP) on theta sweeps can rapidly learn a successor representation, explaining hippocampal place cell properties and spatial prediction mechanisms.

This paper introduces a biologically plausible generative model of the hippocampal formation that uses theta-band oscillations to train a network for spatial and non-spatial memory tasks.

The paper shows place cells fire at progressively earlier ripple phases during replay, indicating phase coding occurs in the ripple band.

This paper develops a deep-learning framework to decode sensory and behavioural variables from wide-band neural data with minimal user input and generalization across stimuli and brain regions.

The paper evaluates three methods for identifying place cells in hippocampal data, finding that the Peak method is most reliable and robust, while different methods identify vastly different populations of place cells.

This study shows that the rate of transitions in the stable place cell population in the hippocampus matches the rate of perceptual change in the environment.

Training a network only to predict its visual input makes head direction, boundary vector, place and egocentric boundary cell responses appear, and lets agents reuse behaviours in unfamiliar environments.

The paper introduces the Tolman-Eichenbaum machine, a computational model that unifies spatial and relational memory by explaining hippocampal and entorhinal cell representations through generalisation principles.

This paper models place and grid cell firing using a successor representation derived from boundary vector cells to explain spatial navigation in the hippocampus.

The paper shows that aligning grid-like neural representations with environmental information axes improves navigation performance and reduces spatial uncertainty.

This study shows that trapezoidal environments distort human spatial memory, similar to how grid cell patterns are affected in rodents, suggesting a link between grid cells and cognitive functions beyond navigation.

This study uses a recurrent neural network to decode rat location from hippocampal recordings, showing improved accuracy over a Bayesian model with flat priors.

The paper identifies neurons in the entorhinal cortex that fire in response to repetitive cues in virtual reality, distinguishing them from true grid cells.

The paper evaluates grid cell identification methods by simulating spatial firing patterns and finds high false positive rates with standard shuffling procedures.

The paper shows that hippocampal replay switches content based on task demands, with task-focused reactivations improving spatial decision accuracy.