Publications
Preprints
| LazyDINO: Fast, Scalable, and Efficiently Amortized Bayesian Inversion via Structure-Exploiting and Surrogate-Drive Measure Transport, Lianghao Cao, Joshua Chen, Michael Brennan, Thomas O’Leary-Roseberry, Youssef Marzouk and Omar Ghattas [preprint] |
| Fast Finite-Sum Optimization via Cyclically-Sampled Hessian Averaging Methods, Thomas O’Leary-Roseberry and Raghu Bollapragada In review at Mathematical Programming [see technical report] |
 |
Inference of Heterogeneous Material Properties via Infinite-Dimensional Integrated DIC Joseph Kirchhoff, Dingcheng Luo, Thomas O’Leary-Roseberry and Omar Ghattas [preprint] |
| Derivative-informed neural operator acceleration of geometric MCMC for infinite-dimensional Bayesian inverse problems Lianghao Cao, Thomas O’Leary-Roseberry, and Omar Ghattas In review at Journal of Machine Learning Research [preprint] |
 |
Efficient PDE-constrained optimization under high-dimensional uncertainty using derivative-informed neural operators Dingcheng Luo, Thomas O’Leary-Roseberry, Peng Chen and Omar Ghattas In review at SIAM Journal on Scientific Computing [preprint] |
Journal Articles
| Scientific Machine Learning: A Symbiosis Brendan Keith, Thomas O’Leary-Roseberry, Benjamin Sanderse, Robert Scheichl and Bart van Bloemen Waanders Foundations of Data Science (2025) |
| SOUPy: Stochastic PDE-constrained optimization under high-dimensional uncertainty in Python Dingcheng Luo, Peng Chen, Thomas O’Leary-Roseberry, Umberto Villa and Omar Ghattas Journal of Open Source Software (2024) [GitHub] [JOSS Paper] |
| Derivative-Informed Neural Operator: An efficient framework for high-dimensional parametric derivative learning Thomas O’Leary-Roseberry, Peng Chen, Umberto Villa and Omar Ghattas Journal of Computational Physics (2024) [JCP paper] [preprint] |
| Residual-based error correction of neural operator accelerated infinite-dimensional Bayesian inverse problems Lianghao Cao, Thomas O’Leary-Roseberry, Prashant Jha, J. Tinsley Oden and Omar Ghattas Journal of Computational Physics (2023) [JCP paper] [preprint] |
| Large-scale Bayesian optimal experimental design with derivative-informed projected neural networks Keyi Wu, Thomas O’Leary-Roseberry, Peng Chen and Omar Ghattas Journal of Scientific Computing (2023) [JSC paper] [preprint] |
| Learning high-dimensional parametric maps via reduced basis adaptive residual networks Thomas O’Leary-Roseberry, Xiaosong Du, Anirban Chaudhuri, Joaquim R. R. A. Martins, Karen Willcox and Omar Ghattas Computer Methods in Applied Mechanics and Engineering (2022) [CMAME paper] [preprint] |
| Derivative-informed projected neural networks for high-dimensional parametric maps governed by PDEs Thomas O’Leary-Roseberry, Umberto Villa, Peng Chen and Omar Ghattas Computer Methods in Applied Mechanics and Engineering (2022) [CMAME paper] [preprint] |
Conference Proceedings
| Learning optimal aerodynamic designs through multi-fidelity reduced-dimensional neural networks Xiaosong Du, Joaquim R. R. A. Martins, Thomas O’Leary-Roseberry, Anirban Chaudhuri, Omar Ghattas and Karen Willcox AIAA SciTech Forum 2023 [SciTech paper] |
| Projected Stein Variational Newton: A Fast and Scalable Bayesian Inference Method in High Dimensions Peng Chen, Keyi Wu, Joshua Chen, Thomas O’Leary Roseberry and Omar Ghattas Proceedings of the 33rd Conference on Neural Information Processing Systems (NeurIPS 2019) [NeurIPS paper] |
Technical Reports
| A note on the relationship between PDE-based precision operators and Matérn covariances Umberto Villa and Thomas O’Leary-Roseberry [arXiv] |
PhD Thesis
| Efficient and dimension independent methods for neural network surrogate construction and training PhD Dissertation, The University of Texas at Austin (2020) [link] |