Abdurakhmon Sadiev

Better LMO-based Momentum Methods with Second-Order Information

Bernoulli-LoRA: A Theoretical Framework for Randomized Low-Rank Adaptation

Error Feedback under $(L_0,L_1)$-Smoothness: Normalization and Momentum

Second-order Optimization under Heavy-Tailed Noise: Hessian Clipping and Sample Complexity Limits

Communication-efficient Algorithms Under Generalized Smoothness Assumptions

A Unified Theory of Stochastic Proximal Point Methods without Smoothness