activity
20182020
most citedOn Optimal Transformer Depth for Low-Resource Language Translation

20 citations · 21 across the 3 of their papers we have counts for

collaborators

8 papers

cs.LG2020

A game-theoretic analysis of networked system control for common-pool resource management using multi-agent reinforcement learning

Arnu Pretorius, Scott Cameron, Elan van Biljon +6

Multi-agent reinforcement learning has recently shown great promise as an approach to networked system control. Arguably, one of the most difficult and important tasks for which la…

cs.CL202020 cited

On Optimal Transformer Depth for Low-Resource Language Translation

Elan van Biljon, Arnu Pretorius, Julia Kreutzer

Transformers have shown great promise as an approach to Neural Machine Translation (NMT) for low-resource languages. However, at the same time, transformer models remain difficult…

cs.LG20191 cited

Stabilising priors for robust Bayesian deep learning

Felix McGregor, Arnu Pretorius, Johan du Preez +1

Bayesian neural networks (BNNs) have developed into useful tools for probabilistic modelling due to recent advances in variational inference enabling large scale BNNs. However, BNN…

cs.LG2019

On the expected behaviour of noise regularised deep neural networks as Gaussian processes

Arnu Pretorius, Herman Kamper, Steve Kroon

Recent work has established the equivalence between deep neural networks and Gaussian processes (GPs), resulting in so-called neural network Gaussian processes (NNGPs). The behavio…

stat.ML2019

If dropout limits trainable depth, does critical initialisation still matter? A large-scale statistical analysis on ReLU networks

Arnu Pretorius, Elan van Biljon, Benjamin van Niekerk +6

Recent work in signal propagation theory has shown that dropout limits the depth to which information can propagate through a neural network. In this paper, we investigate the effe…

cs.CL2019

Unsupervised acoustic unit discovery for speech synthesis using discrete latent-variable neural networks

Ryan Eloff, André Nortje, Benjamin van Niekerk +7

For our submission to the ZeroSpeech 2019 challenge, we apply discrete latent-variable neural networks to unlabelled speech and use the discovered units for speech synthesis. Unsup…