đ€ AI Summary
Existing Stein identities are ill-suited for non-Gaussian distributions with bounded supportâsuch as q-Gaussian distributionsâlimiting their applicability in gradient estimation for generative modeling and stochastic optimization. This work presents the first extension of Bonnet- and Price-type theorems to q-Gaussian distributions by introducing an auxiliary companion distribution, yielding a new Stein identity that is both analytically concise and computationally tractable. The resulting gradient estimator preserves structural similarity to its Gaussian counterpart while significantly reducing variance. Empirical evaluations demonstrate its superior performance in Bayesian deep learning and sharpness-aware minimization tasks, thereby broadening both the theoretical foundations and practical applicability of Steinâs method beyond the Gaussian setting.
đ Abstract
Stein's identity is a fundamental tool in machine learning with applications in generative models, stochastic optimization, and other problems involving gradients of expectations under Gaussian distributions. Less attention has been paid to problems with non-Gaussian expectations. Here, we consider the class of bounded-support $q$-Gaussians and derive a new Stein identity leading to gradient estimators which have nearly identical forms to the Gaussian ones, and which are similarly easy to implement. We do this by extending the previous results of Landsman, Vanduffel, and Yao (2013) to prove new Bonnet- and Price-type theorems for q-Gaussians. We also simplify their forms by using escort distributions. Our experiments show that bounded-support distributions can reduce the variance of gradient estimators, which can potentially be useful for Bayesian deep learning and sharpness-aware minimization. Overall, our work simplifies the application of Stein's identity for an important class of non-Gaussian distributions.