SAMPL: The Spreading Activation and Memory PLasticity Model

Authors

Beau Sievers¹, Ida Momennejad²

¹ Department of Psychology, Harvard University, Cambridge, MA 02138

² Department of Biomedical Engineering, Columbia University, New York, NY 10027

About

This repository contains code for running the simulations and analyses described in the paper "SAMPL: The Spreading Activation and Memory PLasticity Model," available at: https://www.biorxiv.org/content/10.1101/778563v1

This repository does NOT contain the word2vec embedding of the Google News Dataset, which is required to run the Anderson, Bjork, and Bjork (1994) and two-agent memory convergence simulations. Before running these simulations, place the Google News embedding file in the /data directory.

Requirements

All code was tested on Mac OS 10.14.6. Library requirements are listed in requirements.txt.

File structure

• /sampl: code implementing the SAMPL model
• /sampl/parameter_search_all.py: code for running parallelized grid search
• /grid_search_results: cached results from grid search (to save time)
• /notebooks: Jupyter notebooks for running simulations and plotting results
• /data: place the word2vec embedding of the Google News dataset here

Citation

Sievers, B., Momennejad, I., (2019, September 23). SAMPL: The Spreading Activation and Memory PLasticity Model. Retrieved from https://www.biorxiv.org/content/10.1101/778563v1

License

This software is made available under a BSD-style license. See the LICENSE file.

Last updated

This document was last updated on September 23, 2019.