Federico Nola
Multiregion constructions of the neutron star equation of state (EoS) introduce freedom in matching low and intermediate density descriptions. This freedom can generate families of formally admissible stellar models whose physical viability must be assessed against astrophysical observations. We investigated whether current neutron star constraints are sensitive not only to the high density continuation of the EoS, but also to the low density matching procedure itself. To this end, we compared two baseline constructions propagated through a common high density extension. In branch ^ the sub-n_1 EoS was obtained by smoothly matching a low density β-equilibrated input EoS to a β-stable chiral effective field theory (̧hiEFT) branch. In branch the low density branch was continued directly up to n_1=0.32 fm -3 and served as a control construction. Above n_1, both branches were extended using the same parameterized continuation and were confronted with direct NICER mass--radius posteriors, a hard lower bound on the maximum mass, and an effective constraint on Λ_1.4. This common continuation was used to isolate the role of the low density matching sector. The posterior-favored predictions of branches ^ and remain strongly overlapping at the observable level. The inferred stellar properties differ only slightly, and the two branches yield very similar continuation parameters, especially n_2 and Γ_1. Nevertheless, the NICER-informed posterior induces a nontrivial constraint on the matching parameters of branch favoring n_m ≃ 0.18 fm -3 and Δ_m ≃ 0.021 ^ fm -3 . Current astrophysical data constrain the shared continuation above n_1, while indirectly restricting the low density matching sector. The matching is not arbitrary: even nearly degenerate mass--radius predictions can lead the posterior to favor matching regions. This highlights the need to constrain modular EoS frameworks observationally, where density regimes admit multiple valid connections.