Open Dataset: Error-correcting dynamics in visual working memory

Working memory is dynamic, changing over time. In this manuscript, we developed a dynamical systems model to estimate the forces acting on items in working memory. These dynamics are estimated from behavior alone. As detailed in Panichello et al, 2019, we found two forces acted on items in memory: a diffusive force caused memories to randomly diffuse away from their original value and a drift force pulled memories towards attractors in mnemonic space. Behavioral modeling approach is detailed in Panichello et al, 2019.

Code for fitting the behavioral model, as well as example data from the manuscript are available on our lab's GitHub repository. If you are interested in the raw behavioral data, please email Tim.