Metraipolis › archive › dc_2a6d1300ef2d
Semi-Amortized Bayesian Inference with Reusable Local Posteriors — arXiv:2610.18945
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.18945
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-15T14:23:09.457Z
- SHA-256 of the archived text
e2c01bf71883820ff15fcf3c264040861e8293ce2694f4a08b901486e165caf0- Stable URL of this page
https://metraipolis.com/documento/dc_2a6d1300ef2d
How to cite
Semi-Amortized Bayesian Inference with Reusable Local Posteriors — arXiv:2610.18945 — arXiv (en), https://arxiv.org/abs/2610.18945 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_2a6d1300ef2d (Metraipolis archive dc_2a6d1300ef2d, sha256 e2c01bf71883820ff15fcf3c264040861e8293ce2694f4a08b901486e165caf0)
BibTeX
@misc{dc2a6d1300ef2d,
title = {Semi-Amortized Bayesian Inference with Reusable Local Posteriors — arXiv:2610.18945},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.18945},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_2a6d1300ef2d; sha256 e2c01bf71883820ff15fcf3c264040861e8293ce2694f4a08b901486e165caf0},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.18945. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Amortized Bayesian inference uses simulations to perform Bayesian estimation when likelihoods are expensive or analytically unavailable. However, learning the full joint posterior is challenging in high-dimensional models with complex dependencies. We introduce a semi-amortized method that divides parameters into smaller blocks, and uses conditional independence to identify the information needed to infer each block. We train neural estimators for blocks whose posterior updates are computationally expensive or frequently repeated. Each estimator learns from simulations of the relevant model components to approximate the block's full conditional posterior, given the relevant observations and the values of the variables on which it depends. A Markov chain then alternates direct draws from these estimators with conventional updates of the remaining variables. Each neural update uses the current conditioning values and requires neither likelihood evaluation nor an acceptance-rejection step. Where the same conditional structure recurs, one estimator can be reused across blocks and datasets, with an output dimension determined by the block rather than the full model. We illustrate the approach in high-dimensional and computationally expensive hierarchical and spatiotemporal models. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno). Categorie arXiv: Machine Learning (stat.ML) ; Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2610.18945 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-15T13:53:35.808Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).