# rcnn-sat
This project contains weights for pre-trained neural networks and behavioural data used in the paper [Recurrent neural networks can explain flexible trading of speed and accuracy in biological vision](https://doi.org/10.1371/journal.pcbi.1008215). Please cite this publication if you use any of the data from this project.
The networks are both recurrent and feedforward networks that were trained on both [ecoset](https://www2.securecms.com/CCNeuro/docs-0/5927d79368ed3feb338a2577.pdf) and Imagenet. The code for the networks is hosted on [this github repository](https://github.com/cjspoerer/rcnn-sat). The repository contains details of how to use the networks.
The behavioural data are described on [this page](https://osf.io/mz9hw/wiki/Behavioural%20data/).