Abstract:Background: The no-core shell model (NCSM) is an ab initio method that solves the nuclear many-body problem by expanding the many-particle wave function into a (typically) harmonic oscillator basis and minimizing the energy to obtain the expansion coefficients. Extensions of the NCSM, such as its coupling with microscopic-cluster basis states, further allow for an ab initio treatment of light-ion nuclear reactions of interest for both astrophysics and nuclear technology applications. A downside of the method is the exponential scaling of the basis size with increasing number of nucleons and excitation quanta, which limits its applicability to mass A≲16 nuclei, except for variants where the basis is further down-selected via some truncation scheme. Purpose: We consider a basis selection method for the NCSM that was first introduced in the context of the large-scale shell model and captures the essential degrees of freedom of the nuclear wave function leading to a favorable complexity scaling for calculations and enabling ab initio reaction calculations in sd -shell nuclei. Methods: The particle configurations within the NCSM basis are ordered based on their contribution to the first moment of the Hamiltonian matrix that results from the projection onto the many-body basis. The truncation scheme then consists in retaining only the lowest-first-moment configurations, which typically contain only few many-body basis states (Slater determinants). As the energy threshold above which configurations are disregarded is increased, the size of the basis becomes an almost-continuous variable, allowing for tunable fidelity in the obtained wave functions. The resulting wave functions can then be used directly in ab initio reaction calculations. Results: We present calculations for Li7 and n+C12 scattering using nucleon-nucleon interactions derived from chiral effective field theory and softened using the similarity renormalization group method. The obtained energy levels invariably demonstrate exponential convergence with the size of the basis, and we find improved convergence in scattering calculations. To demonstrate the possibilities enabled by the approach, we also present a first calculation for the scattering of neutrons from Mg24 . Conclusions: The method presented in this work appears promising for future studies of nuclei with mass A>16 , opening multiple future research directions impacting both nuclear astrophysics and nuclear technology applications. 2 More Received 4 February 2024 Accepted 17 April 2024 ©2024 American Physical Society Physics Subject Headings (PhySH) Research Areas Collective levelsNuclear many-body theoryNuclear reactionsNuclear structure & decays Nuclear Physics

Deep Learning: Extrapolation Tool for Ab Initio Nuclear Theory.

Deep Learning: A Tool for Computational Nuclear Physics

Deep-neural-network approach to solving the<i> ab</i><i> initio</i> nuclear structure problem

Deep-neural-network approach to solving the ab initio nuclear structure problem

Precise neural network predictions of energies and radii from the no-core shell model

Coupled-Cluster Calculations of Infinite Nuclear Matter in the Complete Basis Limit Using Bayesian Machine Learning

Building Surrogate Models of Nuclear Density Functional Theory with Gaussian Processesand Autoencoders

Deep learning on nuclear mass and $α$ decay half-lives

Emulating ab initio computations of infinite nucleonic matter

Deep learning approach to nuclear masses and -decay half-lives

Ab Initio Description of Open-Shell Nuclei: Merging No-Core Shell Model and In-Medium Similarity Renormalization Group

Nuclear binding energies in artificial neural networks

Large Scale and Quantum Accurate Molecular Dynamics Simulation:Liquid Iron under Extreme Condition

Ab Initio No Core Shell Model - Recent Results and Further Prospects

Ab Initionuclear Structure – the Large Sparse Matrix Eigenvalue Problem

Deep learning for nuclear masses in deformed relativistic Hartree-Bogoliubov theory in continuum

Tunable-fidelity wave functions for the ab initio description of scattering and reactions

Ab Initio Calculation of Real Solids Via Neural Network Ansatz

Machine learning approach to pattern recognition in nuclear dynamics from the ab initio symmetry-adapted no-core shell model

Realizing the potential of deep neural network for analyzing neutron star observables and dense matter equation of state