Metraipolis › archive › dc_d8697b46cf7c
Matching Multi-Loop Complexities with a Single Loop: Optimal Optimization Stationarity and Best-Known Game Stationarity in Nonconvex--Concave Minimax Optimization — arXiv:2609.17973
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.17973
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-15T15:07:53.015Z
- SHA-256 of the archived text
5a30d43cd154e36ea4f86e694826d04f7992dc227e30cfea6e54b9ce0765625e- Stable URL of this page
https://metraipolis.com/documento/dc_d8697b46cf7c
How to cite
Matching Multi-Loop Complexities with a Single Loop: Optimal Optimization Stationarity and Best-Known Game Stationarity in Nonconvex--Concave Minimax Optimization — arXiv:2609.17973 — arXiv (en), https://arxiv.org/abs/2609.17973 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_d8697b46cf7c (Metraipolis archive dc_d8697b46cf7c, sha256 5a30d43cd154e36ea4f86e694826d04f7992dc227e30cfea6e54b9ce0765625e)
BibTeX
@misc{dcd8697b46cf7c,
title = {Matching Multi-Loop Complexities with a Single Loop: Optimal Optimization Stationarity and Best-Known Game Stationarity in Nonconvex--Concave Minimax Optimization — arXiv:2609.17973},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.17973},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_d8697b46cf7c; sha256 5a30d43cd154e36ea4f86e694826d04f7992dc227e30cfea6e54b9ce0765625e},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.17973. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
We introduce a new single-loop algorithmic framework for smooth nonconvex--concave minimax optimization. The resulting projected damped extragradient method combines projected extragradient updates, dual momentum, and a moving proximal center. Under both the optimization-stationarity and game-stationarity criteria, our method achieves the best-known complexity among single-loop first-order methods. For optimization stationarity, our method achieves a gradient complexity of $O(L^2D_Y\bar\Delta_0\varepsilon^{-3})$, where $L$ is the gradient Lipschitz constant, $D_Y$ bounds the diameter of the dual feasible set, and $\bar\Delta_0$ is an initialization quantity involving the value-function gap and the initial gradients. Moreover, by incorporating a fixed-center warm-up phase, the complexity can be improved to $O(L^2D_Y\Delta_\phi\varepsilon^{-3})$, up to an additive lower-order cost, where $\Delta_\phi:=\phi(x_0)-\inf_x\phi(x)$. We further establish a lower bound of $\Omega(L^2D_Y\Delta_\phi\varepsilon^{-3})$ for optimization stationarity over projected zero-respecting first-order methods. This lower bound proves that the warm-started version of our algorithm is optimal up to a constant factor for optimization stationarity within this oracle class. For game stationarity, our method achieves $\mathcal{O}\!(L^{3/2}D_Y^{1/2}\Delta_\phi\varepsilon^{-5/2})$ gradient complexity. This matches the best-known complexity of multi-loop first-order methods, thereby establishing the same complexity with a single-loop algorithmic structure. Under dual strong concavity, the proposed framework achieves $O\!(\sqrt{\kappa}\,L\Delta_\phi\varepsilon^{-2})$ leading complexity for both stationarity criteria, where $\kappa=L/\mu$ is the dual condition number, up to an additive initialization cost. The $\varepsilon^{-2}$ accuracy dependence is optimal under fixed regularity and initialization bounds.
Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-17 in cs.LG).
Categorie arXiv: Optimization and Control (math.OC) ; Machine Learning (cs.LG); Machine Learning (stat.ML).
Pagina del lavoro: https://arxiv.org/abs/2609.17973 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-15T14:08:19.223Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-17?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).