The Kudla-Millson lift is a mathematical tool that takes certain modular forms (functions with rich symmetry properties) and produces geometric objects called differential forms on special curved spaces known as orthogonal Shimura varieties. These spaces arise naturally in number theory and geometry, and understanding their shape, particularly their cohomology groups, reveals deep structural information. This paper extends the theory of these lifts to a broader setting, working over totally real number fields rather than just the ordinary rational numbers. Totally real number fields are number systems built by adjoining solutions to polynomials whose roots are all real, and working over them requires more sophisticated machinery.
The main technical achievement is computing the Fourier expansion of these lifted differential forms, meaning expressing them as infinite sums of explicitly described pieces, each labeled by numerical data. To do this, the authors use an adelic framework, which is a modern number-theoretic approach that handles all prime numbers simultaneously, combined with a tool called the Weil representation and a partial Fourier transform. Working out these expansions in the adelic setting over totally real fields is substantially more involved than in classical cases, and the explicit formulas are the core contribution of the paper.
From these Fourier expansions, the authors draw two important consequences. First, they prove that the lifting process is injective under some mild extra conditions, meaning distinct modular forms produce distinct differential forms and no information is lost. Second, they show the lifted forms are square-integrable and harmonic, which are natural analytical properties, and use this to establish lower bounds on the dimensions of cohomology groups of the Shimura varieties. These lower bounds give concrete evidence about the topological complexity of these spaces, which is a central question in modern arithmetic geometry.