A new converse theorem for Borcherds products
Ingmar Metzler · 2025
Read A new converse theorem for Borcherds products by Ingmar Metzler (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2512-20701.txt.
Introduction
Ingmar Metzler proves a new converse theorem for Borcherds products, a class of automorphic forms built from input data on lattices. He also proves an injectivity result for the Kudla-Millson theta lift in the orthogonal case, in greater generality than previously available. Both results rest on a single assumption: that the base lattice admits one hyperbolic split.
Catalog record
| CREATOR | Ingmar Metzler |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 101 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2512-20701
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
A new converse theorem for Borcherds products Ingmar Metzler Abstract . We establish a new converse theorem for Borcherds products. Moreover, the injectivity of the Kudla–Millson theta lift is demonstrated in the O(n,2) case in greater generality than is currently available in the literature. Both results are derived under the assumption of a single hyperbolic split of the base lattice. 1. Introduction In 1998, Richard Borcherds was awarded the Fields Medal in recognition of his pioneering contributions to the field of automorphic forms and mathematical physics [ OR ] . He had introduced vertex operator algebras as well as a multiplicative lift of automorphic forms and employed these tools to prove the Conway–Norton moonshine conjecture. The constructed lift – nowadays referred to as Borcherds lift – maps vector-valued elliptic modular forms to orthogonal automorphic forms with infinite product expansions. These Borcherds products provide a diverse array of examples of orthogonal modular forms. To be more explicit, let (L,\qq) be an even lattice of signature ({m^{+}},2) and \mathbb{D} be the associated Grassmannian which carries a natural complex structure. The natural arithmetic subgroup to consider is the discriminant kernel \Gamma(L)\leq\OG(L) which is a finite index subgroup stabilising the discriminant group of L . On the elliptic side, let \mathcal{L}=L^{\prime}/L and consider holomorphic modular functions f\colon\mathbb{H}\to\mathbb{C}[\mathcal{L}] of weight k\in\mathbb{Z}/2 , transforming with the dual Weil representation \overline{\rho}_{L} which may possess poles at the cusp. This space is referred to as the space of weakly holomorphic modular forms and is denoted by \mathcal{M}_{L^{-},k}^{!} . Such a form f\in\mathcal{M}_{L^{-},k}^{!} admits a Fourier expansion f(\tau)=\sum_{\mu\in L^{\prime}/L}\sum_{m\in\mathbb{Z}-\qq(\mu)}a(\mu,m)e^{2\pi i\tau m}\mathfrak{e}_{\mu} where \mathfrak{e}_{\mu} denotes the standard basis vector of \mathbb{C}[\mathcal{L}] being 1 in the \mu component and 0 elsewhere. Assume f has integral principal part, meaning that a(\mu,m)\in\mathbb{Z} for all m<0 . Then the Borcherds lift associates to f a meromorphic modular form \Psi(z,f) on \mathbb{D} for \Gamma(L) with unitary multiplier system of finite order such that a) the weight of \Psi(\,\cdot\,,f) is given by a(0,0)/2 , b) the divisor of \Psi(\,\cdot\,,f) equals \frac{1}{2}\sum_{\mu\in L^{\prime}/L}\sum_{n>0}a(\mu,-n)Z(\mu,n), where Z(\mu,n)=\sum_{\begin{subarray}{c}\lambda\in L+\mu\\ \qq(\lambda)=n\end{subarray}}\{z\in\mathbb{D}\mid z\perp\lambda\} is the Heegner divisor of discriminant (\mu,n) (cf. Definition 3.2 ), c) the target form \Psi(\,\cdot\,,f) has explicit infinite product expansions. As previously stated, this lift gives rise to a variety of new orthogonal modular forms and also to the monster denominator formula, which is utilised to prove the moonshine conjecture. In [ Bor98 , Pr. 16.10] , Borcherds himself poses the question whether his lift may be reversed. That is, given a meromorphic modular form F for \Gamma(L) whose zeros and poles are supported on special divisors Z(\mu,n) , is there a form f\in\mathcal{M}_{L^{-},k}^{!} such that \Psi(z,f) equals F up to a non-zero constant factor? In this paper we provide the following answer (cf. Theorem 6.9 ). Theorem 1.1 . Assume that L splits a hyperbolic plane and {m^{+}}>3 . Then every meromorphic modular form F with respect to \Gamma(L) whose divisor is a linear combination of special divisors Z(\mu,n) is (up to a non-zero constant factor) the Borcherds lift \Psi(z,f) of a weakly holomorphic modular form f\in\mathcal{M}_{L^{-},1-{m^{+}}/2}^{!} . The case {m^{+}}=3 is treated separately in Section 6 . More specifically, the same result holds when the Witt rank of L equals 1 , while a weaker version is established for Witt rank 2 . The result stated above improves upon the previously strongest known result [ Bru14 , Thm. 1.2] , which either required an additional scaled hyperbolic split or needed a transition to a sublattice of L . The proof of Theorem 1.1 is intricate. As a first step, we invoke a celebrated duality statement by Bruinier [ Bru02 ] rendering the above converse statement equivalent to the injectivity of another theta lift with wide-ranging applications: the Kudla–Millson lift. The Kudla–Millson lift In the 1980s, Kudla and Millson [ KM86 ] introduced special Schwartz forms {\varphi_{\textnormal{KM}}} on the symmetric spaces attached to the classical groups \OG({m^{+}},{m^{-}}) , \UG({m^{+}},{m^{-}}) , and \Sp({m^{+}},{m^{-}}) , taking values in closed differential forms. Their principal objective was to investigate cohomology classes of special cycles by means of a theta correspondence, generalising the celebrated work of Hirzebruch and Zagier [ HZ76 ] on Hilbert modular surfaces. More precisely, in the orthogonal case, Kudla and Millson symmetrised the Schwartz forms {\varphi_{\textnormal{KM}}} over a base lattice L and let the Weil representation of the symplectic group act on {\varphi_{\textnormal{KM}}} in order to obtain a kernel (cf. Definition 5.3 ) {\Theta_{\textnormal{KM}}}(\tau,z),\quad\tau\in\mathbb{H},\quad z\in\mathbb{D} in a symplectic and an orthogonal variable, which transforms automorphically in both variables. These may then be employed as an integral kernel to shift automorphic objects from the symplectic to the orthogonal setting or vice versa. Such an association is referred to as a theta lift . The authors extended their work in [ KM87 ] in great generality; for an introduction, see [ KM90 ] and [ BF04 ] . The applications of the Kudla–Millson theta correspondence range from the study of the cohomology of orthogonal and unitary Shimura varieties [ KM86 ] and Arakelov theory of Shimura varieties [ Kud04 ] over specific counting problems [ EGS23 ] , constructing mock modular forms and higher dimensional error functions [ FK17 ] to proving a converse theorem for Borcherds products [ Bru02 , Bru14 ] . In this paper, we consider the Kudla–Millson lift on the space of vector-valued elliptic modular forms associated to the Weil representation {\rho}_{L} of an even lattice L of signature ({m^{+}},2) . To be more explicit, let k=\rk(L)/2 and \mathcal{S}_{L,k} denote the space of cusp forms for the Weil representation {\rho}_{L} . Then {\Lambda_{\textnormal{KM}}}\colon\mathcal{S}_{L,k}\to\mathcal{H}^{(1,1)}(Y_{L}),\quad f\mapsto\int_{\Mp_{2}(\mathbb{Z})\backslash\mathbb{H}}\langle f,{\Theta_{\textnormal{KM}}}\rangle\Im^{k}\;\textnormal{d}\mu defines a linear map to the space of square-integrable harmonic differential forms of Hodge type (1{,}1) on the variety Y_{L}=\Gamma(L)\backslash\mathbb{D} and is referred to as the Kudla–Millson lift . Here \mu denotes the hyperbolic measure on the upper half-plane \mathbb{H} on which \Mp_{2}(\mathbb{Z}) acts via Möbius transforms. The question of its injectivity already arose in [ KM90 ] and may be used to compute the rational Picard number of the underlying Shimura variety [ Ber+16 ] , as well as to derive properties of cones generated by special cycles [ BM19 ] [ Zuf22 ] . However, the primary application we have in mind is the converse Theorem 1.1 for Borcherds products. There have been multiple results on the injectivity of the Kudla–Millson lift over the past two decades presented in [ Bru02 , Thm. 5.12] , [ BF10 , Cor. 4.11] , [ Bru14 , Thm. 5.3] , [ Zuf24 , Thm. 6.1] , and [ Ste23 , Cor. 7.8] . These advances are based on three fundamentally different methods. While there has been a recent success in unifying the results above by Zuffetti and the author [ MZ26 , Thm. 6.2] by generalising Zuffetti’s method, this has not shed any new light on the converse theorem for Borcherds products. The most potent result for this application remained Theorem 5.8 by Bruinier which provides injectivity under the assumption that the lattice L splits a hyperbolic plane as well as a scaled hyperbolic plane. In the context of this paper, we present a new approach to proving the injectivity of the Kudla–Millson lift by means of computing cycle integrals on the