Mapping Dark-Matter Clusters via Physics-Guided Diffusion Models

The paper presents a fully automated method for reconstructing the surface mass density of galaxy clusters using photometry and gravitational lensing data. It introduces DarkClusters-15k, a benchmark dataset of 15,000 simulated clusters with paired mass and photometry maps across multiple redshifts and simulation frameworks. By training a diffusion prior on this dataset, the authors generate posterior samples constrained by weak- and strong-lensing observables, achieving accurate, physics‑guided reconstructions with well‑calibrated uncertainties in minutes.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Computer Vision.

• diffusion
• benchmarks