A Stability Benchmark of Generative Regularizers for Inverse Problems

ArXi:2605.10076v1 Announce Type: cross Generative (diffusion) priors nstrate remarkable performance in addressing inverse problems in imaging. Yet, for scientific and medical imaging, it is crucial that reconstruction techniques remain stable and reliable under imperfect settings. Typical definitions of stability encompass the notion of ''convergent regularization'', robustness to out-of-distribution data, and to inaccuracies in the forward operator or noise model. We evaluate these properties numerically.