This paper provides a theoretical explanation for why low-rank adaptation (LoRA) is effective in fine-tuning large pretrained models.

Introduces a new convex optimization framework for logistic scalar-on-matrix regression with low-rank and sparse structures.

This paper investigates convergence properties of regularized Nyström subsampling for unsupervised domain adaptation under covariate shift.

This paper proposes a method for handling overparameterized linear regression using early-stopped negative-shifted gradient descent, which allows for smooth filters and mixed-sign capabilities.

VISReg introduces a novel regularization technique that combines variance control with a Sliced-Wasserstein-based sketching objective to stabilize self-supervised learning, achieving state-of-the-art…

This paper proposes a two-stage algorithm, AC-IHT, for high-dimensional regression with contamination, achieving near-optimal estimation and strong oracle property.

The paper introduces SLORR, a simple and stateless framework for in-training low-rank regularization of neural networks using two variants based on the Hoyer sparsity metric and the nuclear norm.

The paper introduces TRACER, a novel regularization framework that uses Weighted Moving Average (WMA) distillation to robustly finetune multimodal models, mitigating catastrophic forgetting and improv…

The paper proposes ITSPACE, a method for optimizing the Bures-Wasserstein objective on covariance matrices using a proximal majorization-minimization approach.

This paper introduces a Gaussian approximation for the finite-sample distribution of ridge regression estimator, allowing for heteroskedasticity and autocorrelation in the data generating process. Two…

This paper develops a framework for identifying and estimating parameters of interest in automatic debiased machine learning using a Riesz representer, which is identified when it uniquely optimizes a…

The paper investigates the performance improvement of transfer learning in high-dimensional regression.

This paper derives deterministic equivalents for the prediction risk of over-parameterized linear regression with degenerate covariance matrices and dependent covariates, and identifies the configurat…