Export citation

Export citation

Choose format for download:

Download Citation

    Constraining axionlike particle mediated dark matter with observational constraints: A statistical and machine learning approach

    Prashant Thakur1,*, Aravind Taridalu2,†, Ishfaq Ahmad Rather3,‡, Tanech Klangburam4,§, and Chakrit Pongkitivanichkul5,∥

    • 1Department of Physics, Yonsei University, Seoul, 03722, South Korea
    • 2Department of Physics, BITS-Pilani, K. K. Birla Goa Campus, Goa 403726, India
    • 3Institut für Theoretische Physik, Goethe Universität, Max-von-Laue-Straße 1, D-60438 Frankfurt am Main, Germany
    • 4Khon Kaen Particle Physics and Cosmology Theory Group (KKPaCT), Khon Kaen University, 123 Mitraphap Road, Khon Kaen, 40002, Thailand
    • 5Department of Physics, Faculty of Science, Khon Kaen University, 123 Mitraphap Road, Khon Kaen, 40002, Thailand

    • *Contact author: prashantthakur1921@gmail.com
    • †Contact author: p2021@goa.bits-pilani.ac.in
    • ‡Contact author: rather@astro.uni-frankfurt.de
    • §Contact author: klangburam.t@gmail.com
    • ∥Contact author: chakpo@kku.ac.th

    Phys. Rev. D 113, 023001 – Published 2 January, 2026

    DOI: https://doi.org/10.1103/j3mm-zjsv

    Abstract

    We present a comprehensive investigation into the phenomenological consequences of axionlike particle (ALP) mediated dark matter (DM) on the structure of neutron stars (NSs). Our analysis is grounded in a well-established relativistic mean-field framework, featuring nonlinear mesonic self-interactions constrained by nuclear physics data and modern astrophysical observations. We systematically explore the DM parameter space, spanning DM particle masses mχ∈[0,1000]  GeV and DM fermi momenta qf∈[0,0.06]  GeV, by generating a vast ensemble of over 30,000 equations of state (EoSs). This analysis is carried out using two representative hadronic EoSs, a stiff one (EoS1) and a soft one (EoS18), with the explicit inclusion of the crustal EoS to properly account for the low-density regime of NSs. Employing a multitiered statistical filtering scheme, combining voting, likelihood, and kernel density estimation scores, we apply stringent constraints from a suite of multimessenger observations, including radio and x-ray pulsars, GW170817, and low-mass compact object HESS J1731-347, revealing that models satisfying the PSR J0614-3329 radius bound inherently comply with the HESS constraints, positioning ALP-mediated DM as a viable candidate for explaining low-mass compact objects while still supporting 2M⊙ NSs. For the stiff EoS, we obtain a lower bound of mχ≳43 GeV, with score-weighted posteriors favoring qf=0.034−0.012+0.020 and a broad allowed DM mass range mχ∈[101,949] GeV (median ∼466  GeV). In contrast, the soft EoS yields no strict lower bound on mχ, although scenarios with simultaneously large mχ and qf are strongly disfavored. We developed a high-precision supervised interpolation model using AutoGluon to infer DM parameters from reconstructed NS mass-radius curves, achieving R2>0.998. Feature-importance analysis indicates that the DM mass mχ is mainly constrained by global shape indicators such as the radius ratio R1.6/R1.4, whereas the Fermi momentum qf is primarily determined by the tidal deformability Λ1.4.

    Physics Subject Headings (PhySH)

    Authorization Required

    We need you to provide your credentials before accessing this content.

    References (Subscription Required)

    Outline

    Information

    Sign In to Your Journals Account

    Filter

    Filter

    Article Lookup

    Enter a citation