Research map / Optimization / Implicit bias and generalization
Asymptotic theory of SGD with a general learning-rate
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Develops an asymptotic (statistical) theory for SGD under general learning-rate schedules, providing a unified framework for quantifying uncertainty of SGD iterates.
Problem
Polynomially decaying step-sizes give attractive statistical guarantees for SGD but converge slowly in practice, leaving a gap between theory and empirical schedules.
Approach
A unified asymptotic framework that characterizes the limiting distribution of SGD iterates under general (not only polynomially decaying) learning-rate schedules, enabling uncertainty quantification.
Closest papers by meaning
- Noise and Fluctuation of Finite Learning Rate Stochastic Gradient Descent (ICML 2021 · similarity 0.93)
- Statistical Guarantees for High-Dimensional Stochastic Gradient Descent (NeurIPS 2025 · similarity 0.92)
- Global Convergence and Stability of Stochastic Gradient Descent (NeurIPS 2022 · similarity 0.92)
- The Global Convergence Time of Stochastic Gradient Descent in Non-Convex Landscapes: Sharp Estimates via Large Deviations (ICML 2025 · similarity 0.92)
- Tighter Convergence Bounds for Shuffled SGD via Primal-Dual Perspective (NeurIPS 2024 · similarity 0.91)
- Fast Last-Iterate Convergence of SGD in the Smooth Interpolation Regime (NeurIPS 2025 · similarity 0.91)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.