Speaker
Description
Abstract: Neutron stars provide a high-density laboratory to test dark matter (DM) through its gravitational imprint on stellar structure and early evolution. In this talk, I present a unified set of results based on two-fluid modeling of cold neutron stars and evolving proto-neutron stars (PNSs), together with Bayesian model selection, multi-messenger constraints, and inverse parameter inference. For fermionic DM with repulsive self-interactions, correlations between DM microphysics and global observables (mass, radius, and tidal deformability) can appear strong when the nuclear sector is fixed, but become significantly weaker once realistic hadronic equation-of-state (EoS) uncertainties are included, indicating that bulk properties alone may not uniquely identify a DM model. Using RMF-based scenarios (standard NL, a stiffened NL--$\sigma$ cut, and a DM-admixed neutron decay anamoly), Bayesian evidence quantifies how combined nuclear and astrophysical datasets rank the competing mechanisms and exposes where specific constraints introduce tension. For axion-like-particle (ALP) mediated DM, I construct a large ensemble of DM-extended EoSs across $(m_\chi, q_f)$ and confront it with multi-messenger data (radio/X-ray pulsars, GW170817, and HESS~J1731$-$347) using likelihood/KDE-based scoring to isolate the viable parameter region. A supervised surrogate model then enables the inverse mapping from reconstructed mass--radius information to DM parameters, showing that the mass--radius curve-shape indicator $R_{1.6}/R_{1.4}$ primarily constrains $m_\chi$, while $\Lambda_{1.4}$ is most sensitive to the DM Fermi momentum $q_f$. Finally, for PNS evolution with non-annihilating DM coupled only through gravity, DM cores can produce compressional heating whereas extended DM halos can cool the baryonic matter, yielding a qualitatively distinct thermal signature testable with supernova neutrino signals and young pulsar cooling curves.