This project investigates schemes to combine bottom-up / feedforward and top-down / feedback information in deep neural networks.
## Publications
[Poster](https://osf.io/ekycj/) at CCN 2019 and corresponding [paper](https://doi.org/10.32470/CCN.2019.1191-0).
## Software
A software framework for training networks with feedback is available [on GitHub](https://github.com/cJarvers/FeedbackNets.jl). It is based on Julia and Flux.