Reproduce “Domain-Adaptive Neural Posterior Estimation for Strong Gravitational Lens Analysis” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Paxson Swierc, Marcos Tamargo-Arizmendi, Aleksandra Ćiprijanović, Brian D. Nord
The paper tests neural posterior estimation (NPE) with unsupervised domain adaptation (UDA) to infer five lens parameters from simulated images with different source and target noise. UDA aligns features using unlabeled target images and maximum mean discrepancy. On DES-like simulations, Einstein-radius residual falls from 0.26 arcsec by approximately 88%, with improved coverage, but real observations remain untested.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.