Research
My research focuses on optimization, statistical learning theory, and large language models:
- Stability and generalization of stochastic algorithms
- Stochastic and preconditioned optimization
- Riemannian and low-rank optimization
- Data attribution, interpretability, and efficient inference for LLMs
For a complete and up-to-date citation list, see my Google Scholar profile.
Preprints
Masked Language Flow Models
A Nesterov-style Accelerated Gradient Descent Algorithm for the Symmetric Eigenvalue Problem
Selected publications
On-Average Stability of Multipass Preconditioned SGD and Effective Dimension
Does Stochastic Gradient really succeed for Bandits?
Black-Box Uniform Stability for Non-Euclidean Empirical Risk Minimization
Optimization without retraction on the random generalized Stiefel manifold
Compressed sensing of low-rank plus sparse matrices
Publications
Infeasible Deterministic, Stochastic, and Variance-Reduction Algorithms for Optimization under Orthogonality Constraints
An Alternating Minimization Algorithm with Trajectory for Direct Exoplanet Detection — The AMAT Algorithm
A Riemannian Proximal Newton Method
Direct Exoplanet Detection Using L1 Norm Low-Rank Approximation
Low-rank plus sparse trajectory decomposition for direct exoplanet imaging
Likelihood ratio map for direct exoplanet detection
Optimization flows landing on the Stiefel manifold
Matrix rigidity and the ill-posedness of Robust PCA and matrix completion
Multispectral snapshot demosaicing via non-convex matrix completion
A computational approach to hyperspectral imaging for long-range target identification
Differentially expressed microRNAs in lung adenocarcinoma invert effects of copy number aberrations of prognostic genes
Thesis
On low-rank plus sparse matrix sensing