Probing the Dark Sector with AI/ML and Quantum Sensing

This award funds the research activities of Professor Doojin Kim at the University of South Dakota. One of the most important open questions in modern physics is the nature of dark matter, the invisible form of matter that appears to make up a large fraction of the matter in the universe. Although many experiments have searched for dark matter, no non-gravitational interaction has yet to be confirmed, suggesting that new ideas and search strategies are needed. Under this grant award, Professor Kim will investigate light dark-sector particles and related new physics using high-intensity neutrino experiments, artificial intelligence (AI) and machine learning (ML), and emerging quantum-sensing concepts. Neutrino experiments such as DUNE, SBND, and ICARUS create intense beams and large data sets that can be used not only to study neutrinos, but also to search for new particles that interact only very weakly with ordinary matter. AI and ML will be used as discovery tools to identify subtle signal patterns that may otherwise be hidden in complex experimental environments. The project also explores how quantum sensors based on qubit technologies could detect extremely weak photon signals from dark-sector or cosmological particles. This research advances the national interest by promoting scientific progress in one of its most fundamental directions: the discovery and understanding of new physical laws. The project will also train students and early-career researchers in particle physics, AI, high-performance computing, and quantum technologies, while producing open-source tools and outreach activities that broaden the impact of the work. More technically, Professor Kim will pursue an integrated research program at the intersection of dark-sector phenomenology, neutrino physics, ML, and quantum sensing. The first component will develop theoretical and phenomenological frameworks for MeV-scale new physics at beam-based neutrino experiments, including warped extra-dimensional models with light Kaluza-Klein gravitons, multivariate analyses using timing, energy, angular, and spatial correlations, and exotic three-body decays of charged mesons as production channels for light mediators and dark-sector particles. The second component will develop ML methods for neutrino experiments, including classifiers that distinguish sterile-neutrino, non-standard-interaction, and low-mass dark-matter scenarios; empirical models for charged-meson focusing that reduce reliance on computationally expensive horn simulations; and fast detector-response simulations trained on detailed Monte Carlo samples. These tools will support more efficient sensitivity studies and new-physics searches at current and future neutrino facilities. The third component will extend qubit-based photon-detection strategies to ultralight dark-sector particles and relic neutrino-related signals, including scenarios in which axion-like particles or other hidden-sector states produce narrow photon features. Together, these efforts will provide new theoretical insights, reusable computational tools, and concrete experimental strategies for probing weakly coupled particles beyond the Standard Model. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.