Improving Infinitely Deep Bayesian Neural Networks with Nesterov's Accelerated Gradient Method

RSS feed of new/cross-listed arXiv ML papers (Mar 27, 2026) covering advances across Bayesian continuous-depth models, fairness in regression, Bayesian optimization, probabilistic forecasting, student–teacher estimation, functional tensor decompositions, theory of generalization vs memorization, persistence-based topological optimization, large-scale benchmarking of scRNA-seq imputation methods, and analysis of how unconstrained models learn physical symmetries. Highlights include: Nesterov-accelerated-gradient integration to reduce NFEs and improve convergence in SDE-based Bayesian neuralnets