Mihaela Rosca

27 papers A* 6Journal 19Unranked 2
YearRankTypeTitle / Venue / Authors
2024 J jnl
Corridor Geometry in Gradient-Based Optimization.
CoRR
Benoit Dherin, Mihaela Rosca
2024 J jnl
Neural Compression of Atmospheric States.
CoRR
Piotr Mirowski, David Warde-Farley, Mihaela Rosca, Matthew Koichi Grimes, Yana Hasson, Hyunjik Kim, Mélanie Rey, Simon Osindero, Suman V. Ravuri, Shakir Mohamed
2023 J jnl
Implicit regularisation in stochastic gradient descent: from single-objective to two-player games.
CoRR
Mihaela Rosca, Marc Peter Deisenroth
2023 J jnl
Investigating the Edge of Stability Phenomenon in Reinforcement Learning.
CoRR
Rares Iordan, Marc Peter Deisenroth, Mihaela Rosca
2023 J jnl
On a continuous time model of gradient descent dynamics and instability in deep learning.
CoRR
Mihaela Rosca, Yan Wu, Chongli Qin, Benoit Dherin
2023 J jnl
On a continuous time model of gradient descent dynamics and instability in deep learning.
Trans. Mach. Learn. Res.
Mihaela Rosca, Yan Wu, Chongli Qin, Benoit Dherin
2022 A* conf
Why neural networks find simple solutions: The many regularizers of geometric complexity.
NeurIPS
Benoit Dherin, Michael Munn, Mihaela Rosca, David Barrett
2022 J jnl
Why neural networks find simple solutions: the many regularizers of geometric complexity.
CoRR
Benoit Dherin, Michael Munn, Mihaela Rosca, David G. T. Barrett
2021 A* conf
Discretization Drift in Two-Player Games.
ICML
Mihaela Rosca, Yan Wu, Benoit Dherin, David Barrett
2021 J jnl
Discretization Drift in Two-Player Games.
CoRR
Mihaela Rosca, Yan Wu, Benoit Dherin, David G. T. Barrett
2021 A* conf
Spectral Normalisation for Deep Reinforcement Learning: An Optimisation Perspective.
ICML
Florin Gogianu, Tudor Berariu, Mihaela Rosca, Claudia Clopath, Lucian Busoniu, Razvan Pascanu
2021 J jnl
Spectral Normalisation for Deep Reinforcement Learning: an Optimisation Perspective.
CoRR
Florin Gogianu, Tudor Berariu, Mihaela Rosca, Claudia Clopath, Lucian Busoniu, Razvan Pascanu
2020 — conf
A case for new neural network smoothness constraints.
ICBINB@NeurIPS
Mihaela Rosca, Theophane Weber, Arthur Gretton, Shakir Mohamed
2020 J jnl
A case for new neural network smoothness constraints.
CoRR
Mihaela Rosca, Theophane Weber, Arthur Gretton, Shakir Mohamed
2020 J jnl
Monte Carlo Gradient Estimation in Machine Learning.
J. Mach. Learn. Res.
Shakir Mohamed, Mihaela Rosca, Michael Figurnov, Andriy Mnih
2019 A* conf
Deep Compressed Sensing.
ICML
Yan Wu, Mihaela Rosca, Timothy P. Lillicrap
2019 J jnl
Deep Compressed Sensing.
CoRR
Yan Wu, Mihaela Rosca, Timothy P. Lillicrap
2019 J jnl
Monte Carlo Gradient Estimation in Machine Learning.
CoRR
Shakir Mohamed, Mihaela Rosca, Michael Figurnov, Andriy Mnih
2019 A* conf
Training Language GANs from Scratch.
NeurIPS
Cyprien de Masson d'Autume, Shakir Mohamed, Mihaela Rosca, Jack W. Rae
2019 J jnl
Training language GANs from Scratch.
CoRR
Cyprien de Masson d'Autume, Mihaela Rosca, Jack W. Rae, Shakir Mohamed
2018 J jnl
Distribution Matching in Variational Inference.
CoRR
Mihaela Rosca, Balaji Lakshminarayanan, Shakir Mohamed
2018 A* conf
Learning Implicit Generative Models with the Method of Learned Moments.
ICML
Suman V. Ravuri, Shakir Mohamed, Mihaela Rosca, Oriol Vinyals
2018 J jnl
Learning Implicit Generative Models with the Method of Learned Moments.
CoRR
Suman V. Ravuri, Shakir Mohamed, Mihaela Rosca, Oriol Vinyals
2018 — conf
Many Paths to Equilibrium: GANs Do Not Need to Decrease a Divergence At Every Step.
ICLR (Poster)
William Fedus, Mihaela Rosca, Balaji Lakshminarayanan, Andrew M. Dai, Shakir Mohamed, Ian J. Goodfellow
2017 J jnl
Many Paths to Equilibrium: GANs Do Not Need to Decrease a Divergence At Every Step.
CoRR
William Fedus, Mihaela Rosca, Balaji Lakshminarayanan, Andrew M. Dai, Shakir Mohamed, Ian J. Goodfellow
2017 J jnl
Variational Approaches for Auto-Encoding Generative Adversarial Networks.
CoRR
Mihaela Rosca, Balaji Lakshminarayanan, David Warde-Farley, Shakir Mohamed
2016 J jnl
Sequence-to-sequence neural network models for transliteration.
CoRR
Mihaela Rosca, Thomas M. Breuel