On Stability and Robustness of Diffusion Posterior Sampling for Bayesian Inverse Problems
Yiming Yang, Xiaoyuan Cheng, Yi He, Kaiyu Li, Wenxuan Yuan, Zhuo Sun † · 2026
Read On Stability and Robustness of Diffusion Posterior Sampling for Bayesian Inverse Problems by Yiming Yang, Xiaoyuan Cheng, Yi He, Kaiyu Li, Wenxuan Yuan, Zhuo Sun † (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-2602-02045.txt.
Description
Yiming Yang and colleagues study diffusion sampling for Bayesian inverse problems, where such models serve as learned priors. Wrong odds are common and degrade results under outliers. They map the induced drift and prove solvers hold for linear problems, showing weak spots under outliers. To address this, they propose robust diffusion posterior sampling, proven outlier-robust for linear problems and works with gradient samplers. Tests on real inverse problems and natural images show power and strength, with steady gains under outliers.
Catalog record
| CREATOR | Yiming Yang, Xiaoyuan Cheng, Yi He, Kaiyu Li, Wenxuan Yuan, Zhuo Sun † |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 100640 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602.02045