{"id":"arxiv/2608.13915","url":"https://arxiv.org/abs/2608.13915","tags":["math.OC","eess.SY"],"score":0,"title":"First-Order Optimization as Minimum-Time Control","domain":"math.OC","source":"arxiv","authors":["Liraz Mudrik","Isaac Kaminer","Pramod P. Khargonekar"],"reviews":0,"subject":"math.OC","summary":"We formulate first-order optimization as a minimum-time control problem. The iterate is the state, the update, a combination of the gradients observed so far, is the control, and the points with gradient norm at most a prescribed tolerance form the target set. For a fixed objective and start, the minimum number of oracle queries needed to reach the target is a value function: it measures the complexity of the instance, not the worst case over a class. On a strongly convex quadratic, the conjugate gradient iterates emerge from the discrete Pontryagin conditions, and the value is a…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-08-14","reactions":{}}