Research in Elementary Particle Theory

This award funds the research activities of Professors James Halverson, Fabian Ruehle, and Tomasz Taylor at Northeastern University. String theory is a candidate for a unified description of nature that joins Einstein's theory of gravity with the quantum physics of elementary particles. Because the theory governs energies far beyond the reach of any accelerator, testing it requires connecting its mathematical structure to what is observed in the universe at large. Under this award, the three researchers will study these connections and, in doing so, will confront problems whose difficulty has made them a natural proving ground for modern artificial intelligence (AI) and machine learning (ML). A central theme of the funded work is the two-way exchange between fundamental physics and AI: hard physical problems drive the development of new computational methods, while advances in AI make previously intractable calculations possible, including calculations that touch on quantum information science. This project serves the national interest by promoting the progress of science at the frontier of fundamental physics and by advancing the artificial intelligence and quantum information methods that increasingly underpin the nation's scientific and technological enterprise. The broader impacts of the award are substantial. The PIs will train graduate students in a combination of theoretical physics and data science skills that are highly valued across the research economy, and they will participate in the Physics TheoryNet program, which brings working physicists into Boston-area high school classrooms and reaches roughly one thousand students each year. They will further disseminate their results through public lectures, widely followed online seminar series on physics and machine learning, and the release of open-source software. More technically, under this award, the PIs will pursue an interconnected program spanning the string landscape, quantum field theory, and quantum gravity, unified by an emphasis on machine learning and quantum information science. Professor Halverson will study the explicit string axiverse, computing axion masses and couplings in F-theory ensembles containing up to hundreds of thousands of axions and determining the associated dark matter abundances, the fate of the QCD axion quality problem, and the viability of cosmological stasis; he will also analyze the dark glueball problem in concrete compactifications and extend a correspondence between Euclidean field theory and neural networks, studying the origin of locality and constructing interacting supersymmetric theories from Grassmann-valued networks. Professor Ruehle will apply AI and mathematics to the structure of Calabi-Yau moduli spaces, using neural network approximations to Calabi-Yau metrics from his cymetric package to construct back-reacted flux vacua, and, jointly with Halverson, determine Kaluza-Klein scales; he will further develop a volume-based measure for vacuum selection, find invariant function bases for prepotentials with Coxeter symmetries, apply persistent homology to torsion cohomology and special Lagrangian cycles, and use AI techniques to study quantum circuit complexity and its conjectured relation to wormhole geometry. Professor Taylor will develop the theory of particle scattering in de Sitter spacetime, constructing a unitary S-matrix and determining how spacetime curvature imprints on soft photon, gluon, and graviton amplitudes, on the cosmological deformation of the corresponding symmetry algebra, on the connection between de Sitter amplitudes and cosmological correlators, and on neutrino physics. Together these efforts will sharpen the confrontation between string theory and cosmological data while producing broadly applicable advances in the use of artificial intelligence for exact scientific computation. 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.