{"id":"arxiv/2610.41324","url":"https://arxiv.org/abs/2610.41324","tags":["math.OC"],"score":0,"title":"RapidDOC: Riccati-Preconditioned Implicit Differentiation for Differentiable Optimal Control under Operator Mismatch","domain":"math.OC","source":"arxiv","authors":["Xiangda Zhu","Yuankun Chen","Zifei Nie","Yi Chang"],"reviews":0,"subject":"math.OC","summary":"Differentiable optimal control embeds constrained optimization problems in data-driven pipelines by differentiating their optimality conditions and propagating task-loss gradients to model or objective parameters. A key observation is that forward optimization and backward sensitivity computation involve related stage-structured linear systems, making reuse of forward factorization information natural. However, direct reuse is generally invalid, because differences in linearization, barrier terms, regularization, and curvature can create a mismatch between the retained forward operator and…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-10-30","reactions":{}}