The study re-evaluates one-zero and two-zero textures of the neutrino mass matrix using current experimental and cosmological constraints, including oscillation parameters, CMB bounds on the sum of neutrino masses, kinematic limits on effective electron-neutrino mass, and neutrinoless double-beta decay results.

  • For two-zero textures, several structures remain viable under only the CMB bound, with $B$-series textures predicting a Dirac CP phase around $π/2$ and $3π/2$.
  • When stronger CMB+BAO constraints are applied, only $A$-series two-zero textures remain viable.
  • Machine learning techniques, specifically flow matching, were used to analyze one-zero textures, revealing that some structures are already excluded by current data.
  • The allowed one-zero texture structures provide distinct predictions for the sum of neutrino masses ($\sum_i m_i$), effective electron-neutrino mass ($m_{ν_e}^{

m eff}$), effective Majorana mass ($\langle m_{ee}\rangle$), and the Dirac CP phase ($δ_{ m CP}$).

  • The authors discuss how one-zero texture structures can arise from non-invertible selection rules.