Background: Submodular maximization, which characterizes "diminishing marginal returns," is a cornerstone of discrete optimization. It is widely applied in machine learning tasks like feature ...