Science Score: 26.0%
This score indicates how likely this project is to be science-related based on various indicators:
-
○CITATION.cff file
-
✓codemeta.json file
Found codemeta.json file -
✓.zenodo.json file
Found .zenodo.json file -
○DOI references
-
○Academic publication links
-
○Academic email domains
-
○Institutional organization owner
-
○JOSS paper metadata
-
○Scientific vocabulary similarity
Low similarity (7.3%) to scientific vocabulary
Repository
Variational Place Cells
Basic Info
- Host: GitHub
- Owner: bioAI-Oslo
- License: mit
- Language: Jupyter Notebook
- Default Branch: main
- Size: 103 MB
Statistics
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
- Releases: 0
Metadata Files
README.md
Code for reproducing results in Decoding the Cognitive map: Learning place cells and remapping, published in eLife (2024).
To train models and generate results, first generate a dataset, using the create_dataset.ipynb notebook. The dataset contains 500-timestep trajectories visiting six distinct geometries. Network inputs consist of velocities along such trajectories, alongside a time-constant context signal unique to each environment. Labels contain Cartesian coordinates along trajectories. For non-path integrating models, uniformly sampled datasets may also be created, where both labels and inputs are Cartesian coordinates.
Then, create an experiment (i.e., a model) by running model_setup.py. A model name and path can be passed as the first argument to this run. This creates a model directory, wherein a JSON file is created, which specifies model hyperparameters. Edit this file to change e.g. the number of recurrent units.
Finally, train a model by running train_rnn.ipynb (for trajectory data) to train a recurrent network. Subsequent analyses can be found in the notebooks directory. For example, running spatial_representations.ipynb allows for loading a model, running it on a test dataset, and computing ratemaps of unit responses.
Owner
- Name: bioAI
- Login: bioAI-Oslo
- Kind: organization
- Repositories: 11
- Profile: https://github.com/bioAI-Oslo
GitHub Events
Total
- Push event: 3
Last Year
- Push event: 3