Work

Design-Time Learning-Rate Selection and Dispersion-Gated Tempering for Generalised Bayesian Experimental Design

Year: 2026
Type: preprint
Abstract: Reproducible software package, 50-digit python-flint verification scripts, CUDA GPU benchmark suite, and 18-page AMS-LaTeX manuscript for GatedC G-LHW: design-time homoscedasticity and scale certification, multivariate Stein-LHW quadrature calibration, pure-Bayesian AnnealedPredC continuation, and online chi-squared dispersion gating in GBOED.
Language: English
Open Access status: green
OpenAlex ID: W7233552258
Cites:
Cited by:
Related to: