An Improved Last-Iterate Convergence Rate for Anchored Gradient Descent Ascent
Optimization and Control
2026-04-07 v1 Artificial Intelligence
Abstract
We analyze the last-iterate convergence of the Anchored Gradient Descent Ascent algorithm for smooth convex-concave min-max problems. While previous work established a last-iterate rate of for the squared gradient norm, where , it remained an open problem whether the improved exact rate is achievable. In this work, we resolve this question in the affirmative. This result was discovered autonomously by an AI system capable of writing formal proofs in Lean. The Lean proof can be accessed at https://github.com/google-deepmind/formal-conjectures/pull/3675/commits/a13226b49fd3b897f4c409194f3bcbeb96a08515
Cite
@article{arxiv.2604.03782,
title = {An Improved Last-Iterate Convergence Rate for Anchored Gradient Descent Ascent},
author = {Anja Surina and Arun Suggala and George Tsoukalas and Anton Kovsharov and Sergey Shirobokov and Francisco J. R. Ruiz and Pushmeet Kohli and Swarat Chaudhuri},
journal= {arXiv preprint arXiv:2604.03782},
year = {2026}
}