Preprint on 29 quadruply lensed quasars reports Bayes factors 6:1–24:1 favoring core-collapsed halos over CDM

The study models the collective lensing perturbations from low-mass halos and subhalos across the sample, comparing predicted abundances under cold dark matter versus SIDM scenarios that allow core collapse. Authors use a population-level inference framework that incorporates redshift evolution and halo mass functions, then marginalize over uncertainties in globular cluster populations and subhalo internal structure. The data require either an excess of core-collapsed objects or an implausibly high globular cluster count plus more subhalos than N-body simulations predict.

This result connects to earlier hints from strong-lensing flux-ratio anomalies and dwarf-galaxy rotation curves that have long suggested deviations from collisionless CDM on small scales. If core collapse is confirmed, it would tighten constraints on the dark-matter self-interaction cross-section per unit mass to roughly 1 cm² g⁻¹, a regime already under tension with cluster-scale observations. The analysis flags that the current sample size and modeling assumptions leave room for systematic bias from unaccounted baryonic structures.

Next steps include expanding the quadruply imaged quasar sample with JWST and upcoming ELT observations, plus direct comparison against hydrodynamical SIDM simulations that include baryons. Independent tests via stellar-stream gaps or strong-lensing time-delay anomalies could falsify the core-collapse interpretation within three years if the excess signal disappears.