The paper introduces Sources of Uncertainty (SoU) inference, a framework that extends Bayesian inference by allowing heterogeneous confidence to be encoded across model components. This selective treatment controls the uncertainty retained in each component and its propagation, yielding Empirical Bayes and profiling-type inference as special cases. The authors also present a general method for inducing sparsity, demonstrating its application in linear and logistic regression and in Bayesian neural networks for self‑pruning.
arXiv:2406. 12659v3 Announce Type: replace-cross Abstract: We propose a scalable variational Bayes method for statistical inference for a single or pre-specified low-dimensional subset of the coordinates of a high-dimensional parameter in sparse linear regression.
arXiv:2606. 27269v1 Announce Type: cross Abstract: Reliably quantifying predictive uncertainty is difficult for complex, high-dimensional, or misspecified models.
arXiv:2410. 14843v4 Announce Type: replace-cross Abstract: Vanilla variational inference finds an optimal approximation to the Bayesian posterior distribution, but even the exact Bayesian posterior is often not meaningful under model misspecification.
arXiv:2607. 02182v1 Announce Type: new Abstract: Large language models (LLMs) exhibit remarkable reasoning capabilities, but their task-specific fine-tuning is notoriously plagued by overconfidence, severely hindering trustworthy deployment.
The paper presents a tuning‑free empirical Bayes framework for Bayesian generalized linear models that uses a novel mean‑field variational inference algorithm. By estimating the prior within the VI procedure and optimizing the posterior mean directly, the method reduces optimization complexity and supports scalable solvers like L‑BFGS and stochastic gradient descent. Applied to sparse logistic regression, the approach shows superior predictive performance compared to existing methods.
The paper investigates how well Prior-Data Fitted Networks (PFNs) approximate Bayesian posterior predictive distributions (PPDs). By decomposing deployment error into prior, context, and amortization components, the authors find that PFNs generally recover the target PPD accurately, but residual amortization errors remain in rare settings, leading to disagreement among independently pretrained models. They demonstrate that Bayesian model averaging can consistently reduce this uncertainty and improve predictions, even matching full-size TabICLv2 models.
arXiv:2603. 02159v2 Announce Type: replace-cross Abstract: Instrumental variable (IV) and proximal causal learning (Proxy) methods are central frameworks for causal inference in the presence of unobserved confounding.
arXiv:2606. 14235v1 Announce Type: new Abstract: Variational Inference (VI) is a fundamental inference technique in Bayesian machine learning for approximating complex posterior distributions.
arXiv:2605. 00600v2 Announce Type: replace-cross Abstract: Deep neural networks achieve impressive results across diverse applications, yet their overconfidence on unseen inputs necessitates reliable epistemic uncertainty modeling.
arXiv:2603. 12037v2 Announce Type: replace Abstract: Foundation models based on prior-data fitted networks (PFNs) have shown strong empirical performance in causal inference by framing the task as an in-context learning problem.
arXiv:2303. 04345v2 Announce Type: replace Abstract: Federated learning (FL) is a promising framework that models distributed machine learning while protecting the privacy of clients.
arXiv:2607. 21843v1 Announce Type: cross Abstract: Empirical Bayes (EB) performs simultaneous inference across many related latent variables.