arXiv：cs.LG· Xin Zhao, Nico Scherf, Robert Trampel, Kerrin J. Pine, Nikolaus Weiskopf·· 4 小时前AI 评分37
Spectra：基于扩散模型的 SBI 测试时先验精确分量迁移方法
Spectra: Exact Component Transport for Test-Time Prior Adaptation in Simulation-Based Inference
AI 导读
Spectra 是一种面向基于模拟的推断（SBI）的测试时自适应方法，利用精确的 score 迁移恒等式，在无需额外模拟或训练的情况下，通过冻结的扩散模型闭式获得适配后的 score，支持结构化先验变更。在 6 个 SBI benchmark 上，Spectra 在强先验偏移下以较低在线采样成本实现准确适配，使预训练 SBI 模型能够在测试时纳入更新的先验信息。
正文
Abstract:Simulation-based inference (SBI) has become a powerful approach to Bayesian inference in complex scientific models whose likelihoods are difficult or impossible to evaluate. Amortized SBI learns reusable inference models from simulated data, enabling rapid posterior inference for new observations, and modern generative models have made these models increasingly expressive. However, this reuse is limited to the prior distribution chosen during training, whereas scientific analyses often need revised priors as knowledge accumulates or alternative assumptions are tested. We introduce Spectra, a test-time adaptation method for diffusion-based SBI. Spectra uses an exact score-transport identity to obtain the adapted score from a frozen diffusion model in closed form for structured prior changes, without additional simulation or training. Across six SBI benchmarks, Spectra achieves accurate adaptation under strong prior shifts at low online sampling cost. This enables pretrained SBI models to incorporate updated prior information at test time.
| Comments: | 41 pages, 7 figures |
| Subjects: | Machine Learning (cs.LG); Machine Learning (stat.ML) |
| Cite as: | arXiv:2610.08021 [cs.LG] |
| (or arXiv:2610.08021v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.08021 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Xin Zhao [view email]
[v1]
Tue, 6 Oct 2026 09:14:56 UTC (537 KB)
来源：arXiv：cs.LG · arxiv.org