3 papers
stat.ML2026
TAP Accuracy Below the Fluctuation Scale and Universal Posterior Geometry in Spherical Linear Models
Jingbo Liu, Zhiyuan Yu
We study the Bayes-optimal spherical linear model as the ambient dimension and sample size grow proportionally, under a quantitative Marchenko--Pastur spectral-regularity condition…
math.ST2025
Proof of The TAP Free Energy for High-Dimensional Linear Regression with Spherical Priors at All Temperatures
Zhiyuan Yu, Jingbo Liu
Approximate inference is central to Bayesian learning, with variational inference (VI) providing a scalable framework for posterior approximation. While mean-field VI often fails i…
math.ST2024
Sampling from the Random Linear Model via Stochastic Localization Up to the AMP Threshold
Han Cui, Zhiyuan Yu, Jingbo Liu
Recently, Approximate Message Passing (AMP) has been integrated with stochastic localization (diffusion model) by providing a computationally efficient estimator of the posterior m…