Research map / Optimization / Implicit bias and generalization
Last iterate convergence of SGD for Least-Squares in the Interpolation regime.
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Strengthen the fourth-moment and attainable assumptions to log-regularity assumptions | Constant-step-size SGD, last iterate · Noiseless least-squares | expected risk convergence rate: O(ln(T)/T) -> O(1/T) | Helped |
| Strengthen the assumptions to capacity and source conditions | Constant-step-size SGD, last iterate · Noiseless least-squares | expected risk convergence rate: 1/T -> 1/T^{1+α∧β} | Helped |
Closest papers by meaning
- Fast Last-Iterate Convergence of SGD in the Smooth Interpolation Regime (NeurIPS 2025 · similarity 0.92)
- Revisiting the Last-Iterate Convergence of Stochastic Gradient Methods (ICLR 2024 · similarity 0.90)
- An Analysis of Constant Step Size SGD in the Non-convex Regime: Asymptotic Normality and Bias (NeurIPS 2021 · similarity 0.89)
- Nonasymptotic Analysis of Stochastic Gradient Descent with the Richardson–Romberg Extrapolation (ICLR 2025 · similarity 0.88)
- High-dimensional Limit of SGD for Diagonal Linear Networks (NeurIPS 2026 · similarity 0.87)
- Learning Curves of Stochastic Gradient Descent in Kernel Regression (ICML 2025 · similarity 0.87)
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.