Putting an End to End-to-End
In our paper Putting An End to End-to-End: Gradient-Isolated Learning of Representations (S. Löwe, P. O’Connor, B. S. Veeling), we showed that we can train a neural network without end-to-end backpropagation and achieve competitive performance. For this, we received an Honorable Mention for Outstanding New Directions Paper Award at NeurIPS 2019.
Find out more about this paper in this blog-post by Sindy Löwe.