Integrals of stable envelopes for cotangent bundles to Grassmannians

Matthew Crawford, Pavan Kartik, Reese Lance · 2025

Read Integrals of stable envelopes for cotangent bundles to Grassmannians by Matthew Crawford, Pavan Kartik, Reese Lance (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-2510-21573.txt.

Introduction

Matthew Crawford, Pavan Kartik, and Reese Lance compute integrals of stable envelopes, geometric classes that organize how a torus symmetry acts on the cotangent bundle of a Grassmannian. These classes, introduced by Maulik and Okounkov, sit at a crossroads of geometry and mathematical physics, including three-dimensional mirror symmetry.

The integral they define uses equivariant localization, a method that rebuilds global quantities from data at fixed points of the symmetry. Each integral turns out to be an integer times a power of a formal parameter, and the main result is a combinatorial formula for those integers. In the smallest case the integers are plain binomial coefficients, while for larger Grassmannians they appear to be new. The authors conjecture links to curve counting on the mirror-dual space and to richer quiver and bow varieties.

Catalog record

CREATORMatthew Crawford, Pavan Kartik, Reese Lance
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE85 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2510-21573

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Integrals of stable envelopes for cotangent bundles to Grassmannians Matthew Crawford, Pavan Kartik, Reese Lance Abstract We consider cohomological stable envelopes for a natural torus action 𝖳 \mathsf{T} on X = T ∗ ​ G ​ r ​ ( k , n ) X=T^{*}Gr(k,n) , introduced by Maulik-Okounkov. We define the ℂ ℏ ∗ \C^{*}_{\hbar} -equivariant integral of the stable envelope using equivariant localization over the subtorus ℂ ℏ ∗ ⊂ 𝖳 \C^{*}_{\hbar}\subset\mathsf{T} , and compute the integral as a non-equivariant limit of the localization over the full torus, 𝖳 \mathsf{T} . The integral of such a class is an integer times a power of ℏ \hbar , and the main result of this paper is a combinatorial formula for these integers. In 3d mirror symmetry, these non-equivariant limits are expected to reflect some curve counting phenomena on the 3d mirror dual, X ∨ X^{\vee} . When k = 1 k=1 , we obtain the binomial coefficients, and we study some of the combinatorics of the integers for higher k k , which haven’t appeared in the literature before. We give some conjectures and interpretations on extending this phenomena to type A quiver and bow varieties. 1 Overview 1.1 Stable Envelopes Cohomological stable envelopes were introduced by Maulik-Okounkov [ MO12 ] in order to construct geometric actions of the Yangian algebra on the cohomology of Nakajima quiver varieties. These quiver varieties provide a rich class of symplectic resolutions of singularities, and play a key role in the 3D mirror symmetry conjectures. In this paper we restrict our attention to the most fundamental Nakajima quiver variety, the cotangent bundle to the Grassmannian. If a torus \mathsf{T} acts on a smooth variety, X , cohomological stable envelopes are maps of H_{\mathsf{T}}^{\bullet}(\text{pt}) -modules \stab_{\mathfrak{C}}:H_{\mathsf{T}}^{\bullet}(X^{\mathsf{T}})\to H_{\mathsf{T}}^{\bullet}(X) depending on a choice of chamber, \mathfrak{C} , which is a connected component of the complement of all root hyperplanes in an associated root system. The stable envelope of a torus fixed point, p , is the image of its equivariant cohomology class under this map, \stab_{\mathfrak{C}}(p) . According to the theory of equivariant localization via [ AB84 ] , each \stab_{\mathfrak{C}}(p) class is determined by its restrictions to fixed points. Thus we may think of the map \stab_{\mathfrak{C}} as being encoded in the matrix of restrictions (\stab_{\mathfrak{C}})_{ij}=\stab_{\mathfrak{C}}(p_{i})\big|_{p_{j}} In the enumerative formulation of 3D mirror symmetry, the elliptic stable envelopes conjecturally relate certain enumerative counts on X to enumerative counts on its mirror dual, X^{!} , see [ AO21 ] . When setup appropriately this “relation” is via matrix multiplication and simple algebraic changes of variables. Mathematically these counts are given by vertex functions, equivariant curve counts, which are also of physical interest for the corresponding quiver gauge theories. 1.2 Integration and Localization For the rest of the paper, we set X=T^{*}Gr(k,n) and \mathsf{T}=(\C^{*})^{n}\times\C^{*}_{\hbar} . We define the \mathbb{C}^{*}_{\hbar} -equivariant integral, denoted \int_{X}\Stab(p) by viewing \Stab(p) as a class in H_{\C^{*}_{\hbar}}^{*}(X) , and performing equivariant localizing with respect to the torus \C^{*}_{\hbar} . In this case, the integral is expressed as a sum of rational functions in the parameter \hbar representing the coordinate on \C^{*}_{\hbar} . We prove in Proposition ( 2.1.1 ) that the integral 1 1 1 The scaling factor of \hbar is just a convenience for formulas. \int_{X}\hbar^{k(n-k)}\cdot\stab(p_{I}) is actually an integer. We compute this integral in 2 steps: We use \mathsf{T}- equivariant localization, i.e we compute the sum \sum\limits_{p_{J}\in X^{\mathsf{T}}}\frac{\stab(p)|_{p_{J}}}{e(T_{p_{J}}X)} (1) whose terms are rational functions in a_{1},\dots,a_{n} (the coordinates on (\mathbb{C}^{*})^{n}\subset\mathsf{T} ) and \hbar . We then compute the limit of the quantity in ( 1 ) as \textbf{a}\to 0 , and we prove in Proposition ( 2.1.2 ) that this limit equals \int_{X}\stab(p) . Our study is motivated by the role of stable envelopes in 3d mirror symmetry, and its interaction with vertex functions. Originally, the integrals of stable envelopes as we defined appear as coefficients of expansion of the integral representation of vertex functions for T^{*}Gr(k,n) , see [ SV25 ] and [ SV23 ] . In [ SD20 ] , they compute non-equivariant limits of vertex functions on Nakajima quiver varieties, and conjecture that the result is related to representation-theoretic information on the corresponding 3d mirror dual. In [ DS22 ] , they compute Euler characteristics of stable envelopes for the cotangent bundle to the full flag variety, and relate it to an enumerative object on the 3d mirror dual variety called the index vertex. 1.3 Nonequivariant limits The sums appearing in ( 1 ) are, for example, Example 1.1 . Let X=T^{*}Gr(1,4)=T^{*}\mathbb{P}^{3} , and the corresponding torus \mathsf{T}=(\C^{*})^{4} . There are 4 fixed points, p_{1},\dots,p_{4} . Combining terms in the sum, we have \displaystyle\sum\limits_{p_{j}\in X^{\mathsf{T}}}\frac{\stab(p_{1})|_{p_{j}}}{e(T_{p_{j}}X)} \displaystyle=-\frac{1}{(-\hbar+a_{1}-a_{2})(-\hbar+a_{1}-a_{3})(-\hbar+a_{1}-a_{4})} \displaystyle\sum\limits_{p_{j}\in X^{\mathsf{T}}}\frac{\stab(p_{2})|_{p_{j}}}{e(T_{p_{j}}X)} \displaystyle=\frac{3\hbar^{2}-3\hbar a_{1}-a_{1}^{2}-\hbar a_{2}+2\hbar a_{3}-a_{1}a_{3}+2\hbar a_{4}-a_{1}a_{4}+a_{3}a_{4}}{(\hbar+a_{2}-a_{1})(\hbar+a_{3}-a_{1})(\hbar+a_{3}-a_{2})(\hbar+a_{4}-a_{3})(\hbar+a_{4}-a_{2})} \displaystyle\sum\limits_{p_{j}\in X^{\mathsf{T}}}\frac{\stab(p_{3})|_{p_{j}}}{e(T_{p_{j}}X)} \displaystyle=\frac{3\hbar^{2}-2\hbar a_{1}-2\hbar a_{2}+a_{1}a_{2}+\hbar a_{3}+3\hbar a_{4}-a_{1}a_{4}-a_{2}a_{4}+a_{4}^{2}}{(\hbar+a_{3}-a_{1})(\hbar+a_{3}-a_{2})(\hbar+a_{4}-a_{1})(\hbar+a_{4}-a_{2})(\hbar+a_{4}-a_{3})} \displaystyle\sum\limits_{p_{j}\in X^{\mathsf{T}}}\frac{\stab(p_{4})|_{p_{j}}}{e(T_{p_{j}}X)} \displaystyle=\frac{1}{(\hbar-a_{1}+a_{4})(\hbar+a_{4}-a_{2})(\hbar+a_{4}-a_{3})} The non-equivariant limits, which are the highest power of \hbar in the numerator, are exactly the binomial coefficients (1,3,3,1) (times a power of \hbar) . Therefore we are led to guess that the {n\choose 1}=n integers corresponding to each (\C^{*})^{n-1} -fixed point of T^{*}\mathbb{P}^{n-1} are the binomial coefficients: \int_{T^{*}\mathbb{P}^{n-1}}\stab(p_{i})=\binom{n}{i-1}\cdot\hbar^{n-1} and to also wonder what the generalization to T^{*}Gr(k,n) may be. The answer must be some system of integers which generalize the binomial coefficients in some way, which is interesting independent of the geometric setting. The main content of this paper is to present a proof of a closed form for these integrals for the case of T^{*}Gr(k,n) . Theorem (Theorem 3.1 ) . We give an explicit, combinatorial formula F(p_{I})=\int_{X}\Stab(p_{I}) , in terms of only I,n,k . The formula F is a rational function, and it is not clear that, upon substituting in specific values of I,k,n , it becomes an integer. Nevertheless this must be true, again because it is equal to \int_{X}\Stab(p_{I}) . When setting k=1 , F recovers the binomial coefficients. For higher k we find sequences of integers with interesting combinatorial relations. Many statements we will make about the combinatorics of these integers remain as conjectures, and we pose further questions for which we cannot even make conjectures. 1.4 Combinatorics of integrals The n integers \int_{X}\Stab(p) (ignoring the \hbar factor) obtained from T^{*}Gr(1,n)=T^{*}\mathbb{P}^{n-1} are simply the binomial coefficients, and thus can be arranged into Pascal’s triangle. This is an infinite triangle (2-simplex) where each number in a given layer is the sum of the two numbers in the layer directly above it. We prove that a similar phenomenon occurs for the {n\choose 2} integers arising from integrals over T^{*}Gr(2,n) if we arrange them in a 3-simplex, where the integers in one layer are determined solely by the “nearby” integers in the layer above. This nice recurrence relation, which we name 4-neighbor addition, does have exceptions, which we can fix in an ad-hoc way. We demonstrate two methods to modify the combinatorial problem so that 4-neighbor addition holds without exception. These combinatorial structures may be shadows of nearby geometric problems. These results we present for k=2 demonstrate the complexity of this family of integers. This is especially interesting when they are viewed as a generalization of the binomial coefficients. We also provide an outline of a new conjectural method to find the \mathsf{T} -equivariant integrals of stable envelopes due to A. Varchenko. We

Read the complete text