Rustem Islamov

Non-Euclidian Gradient Descent Occurs at the Edge of Stability

Adaptive Methods Are Preferable in High Privacy Settings: An SDE Perspective

Why Do We Need Warm-up? A Theoretical Perspective

Double Momentum and Error Feedback for Clipping with Fast Rates and Differential Privacy

On the Interaction of Batch Noise, Adaptivity, and Compression, under $(L_0,L_1)$-Smoothness: An SDE Approach

Safe-EF: Error Feedback for Non-smooth Constrained Optimization

Adaptive Methods through the Lens of SDEs: Theoretical Insights on the Role of Noise

Towards Faster Decentralized Stochastic Optimization with Communication Compression

Enhancing Optimizer Stability: Momentum Adaptation of NGN Step-size

Enhancing Optimizer Stability: Momentum Adaptation of The NGN Step-size

Momentum and Error Feedback for Clipping with Fast Rates and Differential Privacy