Comparing Fixed and Adaptive Computation Time for Recurrent Neural Networks

March 21, 2018 · Entered Twilight · 🏛 International Conference on Learning Representations

"No code URL or promise found in abstract"
"Derived repo from GitHub Pages (backfill)"

Evidence collected by the PWNC Scanner

Repo contents: .gitignore, LICENSE, README.md, Thesis.pdf, fojo-2018-iclrw-repeatrnn.pdf, src

Authors Daniel Fojo, Víctor Campos, Xavier Giro-i-Nieto arXiv ID 1803.08165 Category cs.NE: Neural & Evolutionary Cross-listed cs.LG Citations 4 Venue International Conference on Learning Representations Repository https://github.com/imatge-upc/danifojo-2018-repeatrnn ⭐ 35 Last Checked 12 days ago

Abstract

Adaptive Computation Time for Recurrent Neural Networks (ACT) is one of the most promising architectures for variable computation. ACT adapts to the input sequence by being able to look at each sample more than once, and learn how many times it should do it. In this paper, we compare ACT to Repeat-RNN, a novel architecture based on repeating each sample a fixed number of times. We found surprising results, where Repeat-RNN performs as good as ACT in the selected tasks. Source code in TensorFlow and PyTorch is publicly available at https://imatge-upc.github.io/danifojo-2018-repeatrnn/