Multiproposal Elliptical Slice Sampling
Guillermina Senn, Nathan Glatt-Holtz, Giulia Carigi, Andrew Holbrook, Håkon Tjelmeland · 2026
Read Multiproposal Elliptical Slice Sampling by Guillermina Senn, Nathan Glatt-Holtz, Giulia Carigi, Andrew Holbrook, Håkon Tjelmeland (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-22358.txt.
Description
Bayesian inference with Gaussian priors often relies on elliptical slice sampling, a self-tuning Markov chain Monte Carlo method. Guillermina Senn and four co-authors generalize it to multiproposal sampling. Several draws are made in parallel at each step, and acceptance follows a distance-informed transition matrix that can favor draws far from the current state.
Parallel draws allow larger moves and faster self-tuning at nearly no extra wall-clock cost when likelihoods are expensive, improving mixing. Theory and experiments suggest dimension-robust mixing, suiting the method to Bayesian inverse problems for partial differential equations. Users gain a sampler that explores high-dimensional posterior spreads more boldly without paying more time.
Catalog record
| CREATOR | Guillermina Senn, Nathan Glatt-Holtz, Giulia Carigi, Andrew Holbrook, Håkon Tjelmeland |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 64 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-22358