No article text was recorded for this version.
last-iterate convergence gap
Version 2 of 2160 papers168 conceptsfrozen 2026-10-04
Papers in this version
- Blog: Open Problem: Is Averaging Needed for Strongly Convex Stochastic Gradient Descent?
- Blog: Making the Last Iterate of SGD Information Theoretically Optimal
- Blog: Convergence rate analysis of several splitting schemes
- Blog: On Almost Sure Convergence Rates of Stochastic Gradient Methods
- Blog: Faster Margin Maximization Rates for Generic Optimization Methods
- Blog: Increasing Iterate Averaging for Solving Saddle-Point Problems
- Blog: Making Gradient Descent Optimal for Strongly Convex Stochastic Optimization
- Blog: Convergence of the Iterates in Mirror Descent Methods
- Blog: Nonasymptotic Bounds for Stochastic Optimization With Biased Noisy Gradient Oracles
- Blog: Improved Last-Iterate Convergence of Shuffling Gradient Methods for Nonsmooth Convex Optimization
- Blog: Demystifying the Myths and Legends of Nonconvex Convergence of SGD
- Blog: A Fast Optimistic Method for Monotone Variational Inequalities
- Blog: A Dynamical System View of Langevin-Based Non-Convex Sampling
- Blog: New Tight Bounds for SGD without Variance Assumption: A Computer-Aided Lyapunov Analysis
- Blog: Beyond Minimax Rates in Group Distributionally Robust Optimization via a Novel Notion of Sparsity
- Blog: The Role of Momentum Parameters in the Optimal Convergence of Adaptive Polyak's Heavy-ball Methods
- Blog: Beyond Least-Squares: Fast Rates for Regularized Empirical Risk Minimization through Self-Concordance
- Blog: Making the Last Iterate of SGD Information Theoretically Optimal
- Blog: Preconditioned primal-dual dynamics in convex optimization: non-ergodic convergence rates
- Blog: The surprising agreement between convex optimization theory and learning-rate scheduling for large model training
- Blog: Subgradient Gliding Method for Nonsmooth Convex Optimization
- Blog: A Variational Perspective on Accelerated Methods in Optimization
- Blog: First-Order Methods with Increasing Iterate Averaging for Solving Saddle-Point Problems
- Blog: Almost Sure Last Iterate Convergence of Sharpness-Aware Minimization
- Blog: Extragradient Method: O(1/K) Last-Iterate Convergence for Monotone Variational Inequalities and Connections With Cocoercivity
- Blog: Last iterate convergence of SGD for Least-Squares in the Interpolation regime
- Blog: Faster and Non-ergodic O(1/K) Stochastic Alternating Direction Method of Multipliers
- Blog: The Step Decay Schedule: A Near Optimal, Geometrically Decaying Learning Rate Procedure
- Blog: Making the Last Iterate of SGD Information Theoretically Optimal
- Blog: Adam with model exponential moving average is effective for nonconvex optimization
- Blog: Parameter-free projected gradient descent
- Blog: Near-optimal Local Convergence of Alternating Gradient Descent-Ascent for Minimax Optimization
- Blog: COMAL: A Convergent Meta-Algorithm for Aligning LLMs with General Preferences
- Blog: Accelerated Single-Call Methods for Constrained Min-Max Optimization
- Blog: Last Iterate is Slower than Averaged Iterate in Smooth Convex-Concave Saddle Point Problems
- Blog: Improved Last-Iterate Convergence of Shuffling Gradient Methods for Nonsmooth Convex Optimization
- Blog: On the Rate of Gaussian Approximation for Linear Regression Problems
- Blog: Langevin Monte Carlo without Smoothness
- Blog: Revisiting the Last-Iterate Convergence of Stochastic Gradient Methods
- Blog: Fast and Strong Convergence of Online Learning Algorithms
- Blog: Last-Iterate Convergence for Generalized Frank-Wolfe in Monotone Variational Inequalities
- Blog: Gradient Descent Averaging and Primal-dual Averaging for Strongly Convex Optimization
- Blog: Tight Nonparametric Convergence Rates for Stochastic Gradient Descent under the Noiseless Linear Model
- Blog: Extragradient Method: O(1/K) Last-Iterate Convergence for Monotone Variational Inequalities and Connections With Cocoercivity
- Blog: Last Iterate is Slower than Averaged Iterate in Smooth Convex-Concave Saddle Point Problems
- Blog: Accelerated Single-Call Methods for Constrained Min-Max Optimization
- Blog: Tight Last-Iterate Convergence of the Extragradient Method for Constrained Monotone Variational Inequalities
- Blog: The Step Decay Schedule: A Near Optimal, Geometrically Decaying Learning Rate Procedure For Least Squares
- Blog: Provable Last-Iterate Convergence for Multi-Objective Safe LLM Alignment via Optimistic Primal-Dual
- Blog: Optimal first-order methods for convex functions with a quadratic upper bound
- Blog: Langevin Monte Carlo without smoothness
- Blog: On the Last Iterate Convergence of Momentum Methods
- Blog: Tight Nonparametric Convergence Rates for Stochastic Gradient Descent under the Noiseless Linear Model
- Blog: On Markov Chain Gradient Descent
- Blog: Momentum-Based Variance Reduction in Non-Convex SGD
- Blog: Negative Stepsizes Make Gradient-Descent-Ascent Converge
- Blog: Parameter-free Stochastic Optimization of Variationally Coherent Functions
- Blog: Convergence of Mean-Field Langevin Stochastic Descent-Ascent for Distributional Minimax Optimization
- Blog: The Road Less Scheduled
- Blog: The Rate of Convergence of Bregman Proximal Methods: Local Geometry Versus Regularity Versus Sharpness
Showing 60 of 160 papers.
History & citation
Current · Permanent link
- last iterate suboptimality split from this direction (50% overlap)
- last-iterate convergence (last-iterate convergence guarantee) split from this direction (54% overlap)
Cite this direction
Latest version (a living page):
@misc{lacuna_last_iterate_convergence_gap_b3505264,
title = {last-iterate convergence gap},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/last-iterate-convergence-gap/rdr_b3505264174b4ea496747bd88b5da24e},
note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-21},
}This exact version (version 2, frozen text):
@misc{lacuna_last_iterate_convergence_gap_b3505264_v2,
title = {last-iterate convergence gap},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/last-iterate-convergence-gap/rdr_b3505264174b4ea496747bd88b5da24e/v/2},
note = {Version 2, frozen 2026-10-04. Accessed 2026-11-21},
}