Physics-based phenomenological modeling of binary black hole hierarchical formation 2: Autodifferentiable functional inference of hierarchical compact-binary populations
Year: 2026
Type: preprint
Abstract: The gravitational-wave (GW) census contains mass and spin structure consistent with contributions from black holes assembled through repeated mergers in dense environments. Connecting that structure to formation physics requires models that are both physically interpretable and tractable within population inference. We construct an autodifferentiable, physics-based phenomenological model in which each dense environment is represented by a coagulation response and a population of such environments produces an observable merger-rate density. Embedded in the gwkokab Poisson-likelihood framework, this model enables joint inference of natal-population and interaction parameters from the GW census. Applied to GWTC-5.0, the framework shows why simple pairwise coagulation models struggle to reproduce the observed high-mass, comparable-mass population and tests alternative interaction structures against the data, while retaining an explicitly modeled natal component.
Source: arXiv (Cornell University)
Authors: O'Shaughnessy, R., Zeeshan, M., Qazalbash, M.
Institution: Rochester Institute of Technology
Language: English
Subfield: Astronomy and Astrophysics
Field: Physics and Astronomy
Domain: Physical Sciences
Open Access status: green
Funder: National Science Foundation
Awards: The Operation and Maintenance of the Laser Interferometer Gravitational Wave Observatory (LIGO), The Construction of the Advanced Laser Interferometer Gravitational Wave Observatory (AdvLIGO), Collaborative Research: WoU-MMA Constraining the nuclear equation of state and population of neutron star mergers through observations of transient and persistent phenomena, WoU-MMA: Inference About Gravitational Wave Sources and Source Populations in the Era of Perpetual Discovery, WoU-MMA: Inference about Gravitational-Wave Sources and Source Populations
OpenAlex ID: W7212040852