Hippocampus as unitary coherent particle filter

  title={Hippocampus as unitary coherent particle filter},
  author={Charles W. Fox and Tony J. Prescott},
  journal={The 2010 International Joint Conference on Neural Networks (IJCNN)},
  • C. Fox, T. Prescott
  • Published 18 July 2010
  • Biology, Computer Science
  • The 2010 International Joint Conference on Neural Networks (IJCNN)
We present a mapping of the hippocampal formation onto a Temporal Restricted Boltzmann Machine [1] based architecture, running a deterministic version of Gibbs sampling, and extended with a lostness detection and recovery circuit modelled on subiculum and septal acetylcholine (ACh). The mapping approximates Bayesian filtering, which infers both auto-associative de-noised percepts and temporal sequences, the latter including sequences of places during navigation. Inference may be viewed as a… 

Figures from this paper

Learning in a Unitary Coherent Hippocampus
This model (UCPF-HC) of the hippocampus as a unitary coherent particle filter, which combines the classical hippocampal roles of associative memory and spatial navigation, using a Bayesian filter framework is extended to include online learning of connections to and from the CA3 region.
Bayesian Integration of Information in Hippocampal Place Cells
This paper proposes that Hippocampal place cells might implement an error correction mechanism by integrating different sources of information in an approximately Bayes-optimal fashion, and compares the predictions of the model with physiological data from rats.
Hippocampal Remapping as Hidden State Inference
It is shown how the hidden state framework can resolve a number of puzzles about the nature of remapping, by formalizing remapping in terms of hidden state inference.
Extending a Hippocampal Model for Navigation Around a Maze Generated from Real-World Data
A previously defined computational memory model, based on hippocampal structure and function, is tested and developed, as a first step towards developing a synthetic model of human ABM SAM, which has functions analogous to that of humanABM.
A computational cognitive framework of spatial memory in brains and robots
Computational cognitive models of spatial memory in navigation space: A review
Computational Modelling of Adult Hippocampal Neurogenesis
A novel computational model of the DG is proposed that incorporates the developmental trajectory of young adult-born dentate granule cells and their developmental trajectory, including changes in synaptic plasticity, connectivity, excitability and lateral inhibition, using a modified version of the restricted boltzmann machine (RBM).
Neural modeling of sequential inferences and learning over episodic memory
Memory and mental time travel in humans and social robots
This contribution describes the use of machine learning techniques—Gaussian process latent variable models—to build a multimodal memory system for the iCub humanoid robot and summarizes results of the deployment of this system for human–robot interaction.


Path integration and the neural basis of the 'cognitive map'
Theoretical studies suggest that the medial entorhinal cortex might perform some of the essential underlying computations by means of a unique, periodic synaptic matrix that could be self-organized in early development through a simple, symmetry-breaking operation.
Computer simulation of hippocampal place cells
A model in which pyramidal cells are the output layer of a competitive-learning, pattern-classification device and patterns of environmental stimuli as viewed by a computerized “rat” from various locations within a simulated environment is presented.
A unified model of spatial and episodic memory
It is shown that a single recurrent attractor network can store both the discrete memories that characterize episodic memory and the continuous representations that characterize physical space in a single network.
Neurons in hippocampal afferent zones of rat striatum parse routes into multi‐pace segments during maze navigation
These hippocampally informed, and in some cases, spatially selective responses support the view of the ventral striatum as an interface between limbic and motor systems, permitting contextual representations to have an impact on fundamental action sequences for goal‐directed behaviour.
A computational model of parallel navigation systems in rodents
A computational model of action selection is presented, in which different experts, implementing locale and taxon strategies, compete in order to select the appropriate behavior for the current task.
Bayesian Computation in Recurrent Neural Circuits
It is shown that a network architecture commonly used to model the cerebral cortex can implement Bayesian inference for an arbitrary hidden Markov model, and a new interpretation of cortical activities in terms of log posterior probabilities of stimuli occurring in the natural world is introduced.
Robust path integration in the entorhinal grid cell system with hippocampal feed-back
This paper sets up the model of a mobile agent equipped with the entorhinal representation of idiothetic and allothetic information and simulated its place learning in a virtual environment, and shows that the continuous feed-back from the integrated hippocampal place representation is able to stabilize the grid cell code.
The pigeon as particle filter
It is suggested that abrupt and unstable learning can be modeled by assuming subjects are conducting inference using sequential Monte Carlo sampling with a small number of samples — one, in the authors' simulations.
Object, space, and object-space representations in the primate hippocampus.
The results show that there are separate as well as combined representations of objects and their locations in space in the primate hippocampus, a property required in an episodic memory system, for which associations between objects and the places where they are seen are prototypical.
Microstructure of a spatial map in the entorhinal cortex
The dorsocaudal medial entorhinal cortex (dMEC) contains a directionally oriented, topographically organized neural map of the spatial environment, whose key unit is the ‘grid cell’, which is activated whenever the animal's position coincides with any vertex of a regular grid of equilateral triangles spanning the surface of the environment.