Quantifying and Modelling Transfer Learning in Mice Between Consecutive Training Stages of a Change Detection Task
I. Magrans de Abril, M. E. Garrett, D. R. Ollerenshaw, P. A. Groblewski, S. Olsen, S. Mihalas
This paper studies transfer learning in mice during a change detection task, finding that performance drops when transitioning to more complex stages and proposing a cognitive model based on deep Q learning with representation remapping.
bioRxiv · Dec 19, 2019 · PreprintAnimal Behavior & Cognition