arXiv：cs.LG· Vincent Pauline, Kirill Tamogashev, Arran Carter, Sanghyeok Choi, Stefan Bauer, Esmeralda S. Whitammer·· 11 小时前
OSIS：基于短时 MCMC 动力学的一步隐式采样摊销推理
Amortised Inference through One-Step Implicit Sampling
AI 导读
研究人员提出一步隐式采样器 OSIS，用短时 MCMC 动力学为隐式建模分布的演化定义局部目标，在渐近极限下该分布成为 MCMC 核的不动点，从而实现对目标分布的采样。OSIS 在采样时比多步摊销采样器和非摊销 MCMC 基线效率高得多，同时保持良好性能。
正文
Abstract:Efficient sampling from Bayesian posterior distributions in high dimensions remains a hard problem, despite advances in theory and algorithms. Multiple steps are usually required to draw unbiased samples, for example, iterations of an MCMC algorithm or integration steps in a diffusion-based amortised sampler. However, scalable one-step amortised inference remains a challenging open problem. In this work, we build upon recent progress in data-driven generative modelling to offer a possible solution for one-step amortised sampling. Our one-step implicit sampler (OSIS) uses short-time MCMC dynamics to define a local target for the evolution of an implicitly modelled distribution. In the asymptotic limit, the modelled distribution is a fixed point of the MCMC kernel and thus samples from the target. We analyse the convergence properties of OSIS and show connections with other generative modelling and sampling algorithms. We study various design choices and show that OSIS performs well compared to multi-step amortised samplers and non-amortised MCMC baselines while being far more efficient at sampling time.
| Subjects: | Machine Learning (cs.LG); Machine Learning (stat.ML) |
| Cite as: | arXiv:2610.12679 [cs.LG] |
| (or arXiv:2610.12679v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.12679 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Vincent Pauline [view email]
[v1]
Thu, 8 Oct 2026 20:01:57 UTC (22,504 KB)
来源：arXiv：cs.LG · arxiv.org