BFGS can converge to non-Clarke-stationary points with positive probability
math.OC
Submitted: 2026-10-25
Updated: 2026-11-04
Comments: 23 pages, 2 figures. Verification code and Lean formalization: https://github.com/LexiaoLai/Nonsmooth-BFGS-Nonstationary
License: http://creativecommons.org/licenses/by-nc-nd/4.0/
The gist: We construct a coercive, locally Lipschitz, semialgebraic function on R squared for which BFGS with Armijo--Wolfe line search converges to a non-Clarke-stationary point from a nonempty open set of…
Abstract
We construct a coercive, locally Lipschitz, semialgebraic function on R squared for which BFGS with Armijo--Wolfe line search converges to a non-Clarke-stationary point from a nonempty open set of initial points and inverse-Hessian estimates. This disproves the almost-sure stationarity assertion in the 2013 challenge of Lewis and Overton. The construction and proof combine analytical arguments with computer-assisted verification of explicit inequalities.
Related papers
- An Accelerated Proximal Gradient Method for Bures-Wasserstein Space
- FOMTools: Development and AI-Assisted Code Generation for First-Order Methods
- Lions and Muons: Optimization via Stochastic Frank-Wolfe under Heavy-Tailed Noise
- Dynamic Service Recommendation with Congestion-Dependent Joining: Near-Optimal and Constant-Factor Approximation Algorithms
- Integrating Learning and Optimization for Open Transport Capacity Pooling under Disruption
- Zeroth-Order Gradient Estimators under Decision-Dependent Noise