Nonzero 𝔫 \mathfrak{n} cohomology of TDLDS representations
Jin Lee · 2025
Read Nonzero 𝔫 \mathfrak{n} cohomology of TDLDS representations by Jin Lee (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-2507-23102.txt.
Introduction
Jin Kunwoo Lee studies a special class of representations of unitary groups, in a solo math paper from two thousand twenty-five. Representations are ways to realize group actions as linear maps. The class at issue sits at a degenerate limit of the discrete series, a family tied to square-summable behavior.
Lee shows one can choose a cohomology group for such a representation that does not vanish. The choice sits at a canonical degree, and the groups obey a duality rule. Twin cases for two group sizes share the same live degree. A linking map for two small cases closes the paper, with a subtle vanishing twist.
Catalog record
| CREATOR | Jin Lee |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 48 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2507-23102
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Nonzero 𝔫 \mathfrak{n} cohomology of TDLDS representations Thanks: The author is partially supported by NSF RTG grant DMS 2231514. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author and do not necessarily reflect the views of the National Science Foundation. Jin Lee Abstract . We show that a totally degenerate limit of discrete series representation admits a choice of n cohomology group that is nonvanishing at a canonically defined degree. We then show that the combinatorial complexes used by Soergel to compute these cohomology groups satisfies Serre duality. We conclude that this produces two n cohomology groups, each for a totally degenerate limit of discrete series of U(n+1) and U(n), which are nonvanishing at the same degree. This suggests Gan Gross Prasad type branching laws for the TDLDS of unitary groups of any rank. 1. Introduction Totally degenerate limits of discrete series (TDLDS) are relatively unexplored but tantalizingly close to discrete series representations, which are well understood in automorphic representation theory. It is expected that they correspond to Galois representations, but currently known methods provide no robust link between them and arithmetic geometry in any degree of generality. It was observed by Carayol that there is a link between TDLDS representations and geometry, in the sense that they are detected in the coherent cohomology of Griffiths Schmid domains with coefficients in a holomorphic line bundle. The domains where the line bundles exhibit these TDLDS representations in their cohomology cannot be Hermitian symmetric, and in fact do not admit any structure as a quasi projective variety [ GRT13 ] . We must therefore study these line bundles in terms of their equivariant structure over a homogeneous complex manifold without the classical tools from algebraic geometry. Our main tool is the cohomology of TDLDS with respect to a Lie algebra \mathfrak{n} . It was shown rather suggestively by Carayol that in certain cases these cohomology groups are non zero. Our goal in this paper is to extend Carayol’s results. Our contribution is to note that there are two canonical, non zero \mathfrak{n} cohomology classes associated to any TDLDS representation V , and the degree of these classes is given by the dimension and codimension of a certain embedding of closed varieties that can be defined in terms of \mathfrak{n} and V . Our main tool is a combinatorial complex constructed by Soergel that computes these \mathfrak{n} cohomology groups without explicit dimensions and degrees in which one may hope to find cohomology in our case of interest. 1.1. Nonvanishing of \mathfrak{n} -chomology groups To state our main result, we need to specify what we mean by \mathfrak{n} cohomology. Let V be a unitary TDLDS representation of a reductive linear algebraic group G over \mathbb{Q} with anisotropic maximal torus T . Let K{}(\mathbb{R}) be a maximal compact subgroup of G(\mathbb{R}) containing T(\mathbb{R}) . Let \mathfrak{g}{}_{\mathbb{R}},\mathfrak{k}_{\mathbb{R}} and \mathfrak{t}_{\mathbb{R}} be the respective Lie algebras of G(\mathbb{R}) , K{}(\mathbb{R}) and T(\mathbb{R}) , and let \mathfrak{g}=\mathfrak{g}{}_{\mathbb{R}}\otimes_{\mathbb{R}}\mathbb{C} and \mathfrak{k}=\mathfrak{k}_{\mathbb{R}}\otimes_{\mathbb{R}}\mathbb{C} and \mathfrak{t}=\mathfrak{t}_{\mathbb{R}}\otimes_{\mathbb{R}}\mathbb{C} be their complexifications. Let \Phi=\Phi(\mathfrak{g},\mathfrak{t})\supset\Phi^{+} be choices of root system and positive roots with respect to \mathfrak{t} and let \rho be the half sum of positive roots. Our choice of maximal compact K{}(\mathbb{R}) also provides us with a notion of compact roots \Phi_{K}=\Phi(\mathfrak{k},\mathfrak{t}) , defined as the root system of \mathfrak{k} with respect to \mathfrak{t} . Write \mathfrak{n}\subset\mathfrak{g} as the direct sum of negative root spaces of \mathfrak{g} . We define the \mathfrak{n} cohomology of the representation V as the Lie algebra cohomology H^{*}(\mathfrak{n},V) . These groups admit an action of \mathfrak{t} induced by the adjoint action on \mathfrak{n} and the action of \mathfrak{g} on V . In this paper we are concerned with the weight spaces H^{*}(\mathfrak{n},V)_{-\lambda} of this action with respect to weights \lambda\in\mathfrak{t}^{\vee} . Our main result is as follows: Theorem 1.1 . Given an integral weight \lambda\in\mathfrak{t}^{\vee} , there exists a choice of limit of discrete series representation V with infinitesimal character \lambda+\rho such that H^{q}(\mathfrak{n},V)_{-\lambda} is nonzero at q=\#(\Phi_{K}\cap\Phi^{+}) Recall that these cohomology groups must satisfy Serre duality, as is shown in [ BW80 , Proposition 1.5] . We therefore prove that the combinatorial complex used to compute these cohomology groups also satisfies Serre duality: Theorem 1.2 . Given the choices in the previous theorem, one has H^{q}(\mathfrak{n},V)_{-\lambda}\simeq H^{\#\Phi^{+}-q}(\mathfrak{n},V^{\vee})_{\lambda+2\rho} However, the proof of this theorem reveals that the Serre duality of H^{*}(\mathfrak{n},V) is compatible with the Serre duality of the \mathfrak{n}\cap\mathfrak{k} cohomology of the K types of V . The main idea of Theorem 1.1 uses the fact that a TDLDS representation can only be nonzero if the simple roots defining it are non-compact. This forces the differentials computed by Soergel to be 0 at the canonical degrees. We then noticed that the degrees in which these cohomology groups occur for quasi split unitary groups suggest a geometric application of the Gan Gross Prasad conjecture. 1.2. Geometrization of the GGP conjecture In section 6 we construct holomorphic line bundles \mathcal{L}_{\lambda_{m}} over a family of compact complex analytic manifolds \mathrm{GS}(\mathrm{U}_{m}) for m=n,n+1 . The resulting cohomology H^{*}(\mathrm{GS}(\mathrm{U_{m}}),\mathcal{L}_{\lambda_{m}}) is given in terms of the \mathfrak{n} cohomology groups discussed above. Motivated by the GGP conjectures, we construct a restriction map \displaystyle H^{*}(\mathrm{GS}(\mathrm{U_{n+1}}),\mathcal{L}_{\lambda_{n+1}})\mathrel{\mathchoice{{}\hbox{$\displaystyle{\meno}$}}{{}\hbox{$\textstyle{\meno}$}}{{}\hbox{$\scriptstyle{\meno}$}}{{}\hbox{$\scriptscriptstyle{\meno}$}}}\mathrel{\mkern-3.0mu}\rightarrow H^{*}(\mathrm{GS}(\mathrm{U_{n}}),\mathcal{L}_{\lambda_{n}}) It is natural to ask whether this map is nonzero. If it is, then the restriction map can be thought of as a geometric manifestation of the period integral occuring in the GGP conjectures. As we explain in section 6 , our work implies that in the Archimedean place there is no cohomological obstruction to the nonvanishing of this period. 1.3. Outline We begin by defining the representations and cohomology groups of interest in section 2 . We then introduce the root data constructions in section 3 . We also included some proofs regarding these tools in Appendix A . Our main results are stated in sections 4 and 5 where we give the nonvanishing result in theorem 4.1 and Serre duality in theorem 5.3 . The specific degrees in which these cohomology groups occur in the case of unitary groups is given in section 6 . We then relate our results to a cohomological obstruction of the nonvanishing of the period integrals occuring in the GGP conjectures. 2. Notation 2.1. Harish Chandra isomorphism for central characters Throughout this section we shall assume all groups are defined over \mathbb{R} . Let G be a reductive linear algebraic group with anisotropic maximal torus T . Thus T(\mathbb{R}) is compact. Let K{}(\mathbb{R})\leq G(\mathbb{R}) be a maximal compact subgroup containing T(\mathbb{R}) and let \mathfrak{t}_{\mathbb{R}}\subset\mathfrak{k}_{\mathbb{R}}\subset\mathfrak{g}{}_{\mathbb{R}} be the corresponding inclusion of Lie algebras. Let \mathfrak{t}\subset\mathfrak{k}\subset\mathfrak{g}=\mathfrak{g}{}_{\mathbb{R}}\otimes_{\mathbb{R}}\mathbb{C} be the respective complexifications which correspond to the inclusion of groups of complex points T(\mathbb{C})\subset K(\mathbb{C})\subset G(\mathbb{C}) . Let \Phi:=\Phi(\mathfrak{g},\mathfrak{t}) be the set of roots, and \Phi_{K} be the set of compact roots. We recall that \Phi_{K} can be defined as the roots that occur in the root space decomposition \mathfrak{k}=\mathfrak{t}\oplus(\bigoplus_{\alpha\in\Phi_{K}}\mathfrak{g}_{\alpha}) of \mathfrak{k} by the action of \mathfrak{t} . The Weyl group shall be denoted W:=W(\mathfrak{g},\mathfrak{t}) , and the compact Weyl group W_{K}\subset W is the subgroup generated by the reflections coming from \Phi_{K} . Let \Phi^{+}\supset\Sigma be a choice of positive roots and simple roots, let \rho be the half sum of positive roots, and