{"id":"arxiv/2609.33630","url":"https://arxiv.org/abs/2609.33630","tags":["math.OC","cs.LG","eess.SY"],"score":0,"title":"lapanda: A Matrix-Free Differentiable Solver for Nonconvex Constrained Optimization Layers","domain":"math.OC","source":"arxiv","authors":["Yuankun Chen","Zifei Nie","Kangyu Lin","Ján Drgoňa","Liang Wu"],"reviews":0,"subject":"math.OC","summary":"Differentiable optimization brings the structural guarantees of mathematical optimization to network pipelines, allowing them to be trained end-to-end. However, its application remains challenging for nonconvex constrained problems, as existing differentiable solvers often suffer from limited modeling expressiveness due to their reliance on specialized problem structures, while also incurring substantial computation time and memory overhead in both the forward and backward passes. To address these challenges, we propose lapanda, a matrix-free differentiable solver for nonconvex optimization…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-09-27","reactions":{}}