research
∙
07/11/2023
Implicit regularisation in stochastic gradient descent: from single-objective to two-player games
Recent years have seen many insights on deep learning optimisation being...
share
research
∙
07/09/2023
Investigating the Edge of Stability Phenomenon in Reinforcement Learning
Recent progress has been made in understanding optimisation dynamics in ...
research
∙
09/27/2022
Why neural networks find simple solutions: the many regularizers of geometric complexity
In many contexts, simpler models are preferable to more complex models a...
research
∙
05/28/2021
Discretization Drift in Two-Player Games
Gradient-based methods for two-player games produce rich dynamics that c...
research
∙
05/11/2021
Spectral Normalisation for Deep Reinforcement Learning: an Optimisation Perspective
Most of the recent deep reinforcement learning advances take an RL-centr...
research
∙
12/14/2020
A case for new neural network smoothness constraints
How sensitive should machine learning models be to input changes? We tac...
research
∙
06/25/2019
Monte Carlo Gradient Estimation in Machine Learning
This paper is a broad and accessible survey of the methods we have at ou...
research
∙
05/23/2019
Training language GANs from Scratch
Generative Adversarial Networks (GANs) enjoy great success at image gene...
research
∙
05/16/2019
Deep Compressed Sensing
Compressed sensing (CS) provides an elegant framework for recovering spa...
research
∙
06/28/2018
Learning Implicit Generative Models with the Method of Learned Moments
We propose a method of moments (MoM) algorithm for training large-scale ...
research
∙
02/19/2018
Distribution Matching in Variational Inference
The difficulties in matching the latent posterior to the prior, balancin...
research
∙
10/23/2017
Many Paths to Equilibrium: GANs Do Not Need to Decrease a Divergence At Every Step
Generative adversarial networks (GANs) are a family of generative models...
research
∙
06/15/2017
Variational Approaches for Auto-Encoding Generative Adversarial Networks
Auto-encoding generative adversarial networks (GANs) combine the standar...
research
∙
10/29/2016