The NilHecke schober

Jasper van de Kreeke · 2025

Read The NilHecke schober by Jasper van de Kreeke (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-2508-21737.txt.

Introduction

Jasper van de Kreeke, writing alone in a 2025 paper, builds a mathematical object called a perverse schober. Schobers are used to describe certain categories that arise in geometry and knot theory, but they are hard to define and harder to construct. His target is a category tied to a space studied in a program that turns knot invariants into richer structures.

The construction is fully algebraic and rests on a simpler family of schobers built by others. Diagram methods, where equations become pictures, help check that the new object meets every defining rule. The author aims for a model that matches the intended category exactly. Such explicit models matter because the objects are often invoked but rarely written down in full.

Catalog record

CREATORJasper van de Kreeke
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE92 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2508-21737

Browse the full library · About the Tower of Babel

Full text

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

The NilHecke schober Jasper van de Kreeke Abstract Perverse schobers can be used to describe Fukaya categories but are hard to axiomatize and construct. In this paper, we give an explicit construction of a perverse schober intended to accurately describe the Fukaya category of the horizontal Hilbert scheme considered by Aganagic et al. in the frame of the knot invariant categorification program. The formalism is based on the notion of A n A_{n} -schobers of Dyckerhoff and Wedrich. The construction is entirely algebraic and we check the schober axioms with the help of diagrammatic calculus. Contents 1 Introduction Perverse schobers are a concept of “sheaf of categories” over a given stratified base space B . In the case of the symmetric product B=\operatorname{Sym}^{n+1}(\mathbb{C}) , a full definition of perverse schober was achieved by Dyckerhoff-Wedrich [ 5 ] . It entails the datum of a cubical diagram of stable \infty -categories satisfying certain axioms: \chi:\{0,1\}^{n}\to\operatorname{St}. The horizontal Hilbert scheme \operatorname{Hilb}^{\operatorname{hor}}_{n}(\mathbb{C}^{2}) is a Liouville manifold and its Fukaya category with potential was studied in [ 3 ] . Up to generation questions, this category is entirely described as \operatorname{Perf}\operatorname{NH}_{n} where \operatorname{NH}_{n} is the NilHecke algebra on n strands. The perfect complexes of NilHecke algebras assemble into a cubical diagram as well: \displaystyle\mathbf{NH}_{n+1}:\{0,1\}^{n}\cong\operatorname{Comp}(n+1) \displaystyle\to\operatorname{St}, \displaystyle(\sigma_{1},\ldots,\sigma_{k}) \displaystyle\mapsto\operatorname{Perf}(\operatorname{NH}_{\sigma_{1}}\otimes_{\mathbb{C}\llbracket\hbar\rrbracket}\ldots\otimes_{\mathbb{C}\llbracket\hbar\rrbracket}\operatorname{NH}_{\sigma_{k}}). In this paper, we check that \mathbf{NH}_{n+1} is a perverse schober on \operatorname{Sym}^{n+1}(\mathbb{C}) in the sense of Dyckerhoff-Wedrich. Perverse schobers The concept of perverse schobers was introduced by Kapranov and Schechtman [ 6 ] as the closest approximation to the notion of perverse sheaf of categories with respect to a given stratification. Their primary use case has been to describe Fukaya categories and similar categories in a way that allows for \infty -theoretical local constructions, effectively making the treatment more ergonomic compared to global methods. The precise notion of perverse schober depends on the given base space and its stratification and is often hard to axiomatize. The reason is that a perverse schober on a given base should precisely decatigorify, upon taking K_{0} , to the corresponding notion of perverse sheaf on the same base space. Accurate notions of perverse schobers have been given so far for marked disks [ 6 ] , marked surfaces [ 4 ] and symmetric powers \operatorname{Sym}^{k+1}(\mathbb{C}) [ 5 ] . Fukaya categories of Coulomb branches Fukaya categories of Coulomb branches have been studied in [ 3 ] , as part of the program to categorify knot invariants via mirror symmetry [ 1 , 2 ] . The result of that paper is an explicit description of the subcategory of \operatorname{Fuk}(\mathcal{M}^{\times}(Q,d),W) given by a certain set of objects T_{\theta} . More precisely, there is a matching of objects T_{\theta} with red and black dot distributions on the line, and a matching of their hom spaces with linear spans of strand diagrams between two configurations. This amounts to the inclusion \operatorname{Perf}(\operatorname{KLRW})\subseteq\operatorname{Fuk}(\mathcal{M}^{\times}(Q,d),W) . Let us focus on the case of (Q,d) being the A_{1} -quiver without framing and with arbitrary dimension d=n+1 . In this case, we have \mathcal{M}^{\times}(Q,d)=\operatorname{Hilb}^{\operatorname{hor}}_{n+1}(\mathbb{C}^{2}) and \operatorname{KLRW}=\operatorname{NH}_{n+1} . The horizontal Hilbert scheme comes with a map \pi:\operatorname{Hilb}^{\operatorname{hor}}_{n+1}(\mathbb{C}^{2})\to\operatorname{Sym}^{n+1}(\mathbb{C}) , which is a fiber bundle over the open locus \operatorname{Sym}_{\neq}^{n+1}(\mathbb{C}) consisting of points with pairwise distinct coordinates. This suggests that \operatorname{Fuk}(\operatorname{Hilb}^{\operatorname{hor}}_{n+1}(\mathbb{C}^{2}),W) is the global sections category of a “sheaf of categories” over \operatorname{Sym}^{n+1}(\mathbb{C}) . The schober perspective makes for the hypothesis that [ 3 ] is in fact the shadow of an equality between two different schobers: the “algebraic” NilHecke schober \mathbf{NH}_{n+1} and a “geometric” horizontal Hilbert scheme schober F_{x}\coloneqq\operatorname{Fuk}(\pi^{-1}(x),W|_{\pi^{-1}(x)}) . The present paper is a first step in that direction. Structure of the paper In section 2 , we recall the definition of A_{n} -schobers according to Dyckerhoff-Wedrich [ 5 ] . In section 3 , we introduce the NilHecke schober \mathbf{NH}_{n+1} and check that it satisfies the A_{n} -schober axioms. In Appendix A , we demonstrate the verification of the schober axioms in two specific examples cases. In Appendix B , we sketch certain cases of the proof which have been left out of the main discussion. 2 Preliminaries on A_{n} -schobers In this section, we recall the definition of A_{n} -schobers according to Dyckerhoff-Wedrich [ 5 ] . We start with the cubical indexing set \{0,1\}^{n} and its bijective correspondence with compositions of n+1 . After that, we provide explicit descriptions of bifactorization cubes and Beck-Chevalley cubes. \chi:\{0,1\}^{n}\to\operatorname{Cat}_{\infty} Q_{\chi}((a,b),(c,d)) \mathrm{BC}\big(Q_{\chi}((a,b),(c,d))\big) T_{(a,b),(c,d)}=\operatorname{fib}\!\big(\mathrm{BC}(\,Q_{\chi}((a,b),(c,d))\,)\big) bifactorization Beck–Chevalley fiber (for each (a,b),(c,d)\in\operatorname{Comp}(n{+}1) ) supposed to be equivalence (if (a,b)=(d,c) ) or vanish (else) The construction of bifactorization cubes and Beck-Chevalley cubes in principle applies to any given cubical diagram (of categories or parallel functors, respectively). However, the definition of A_{n} -schobers references a chained and very specific subset of these constructions. The definition starts from the datum of one cubical diagram \chi:\{0,1\}^{n}\to\operatorname{Cat}_{\infty} . Consequently, for each pair of compositions (a,b),(c,d)\in\operatorname{Comp}(n+1) one constructs the bifactorization cube Q_{\chi}((a,b),(c,d)) . Then one constructs the Beck-Chevalley cube \operatorname{BC}(Q_{\chi}((a,b),(c,d))) . Finally, one takes the fiber T_{(a,b),(c,d)}\coloneqq\operatorname{fib}(\operatorname{BC}(Q_{\chi}((a,b),(c,d)))) . The definition of A_{n} -schobers entails that these fibers T_{(a,b),(c,d)} for every (a,b),(c,d)\in\operatorname{Comp}(n+1) satisfy certain conditions. In section 2.1 , we recall notation for cubical diagrams. In section 2.2 , we recall the notion of bifactorization cubes. In section 2.3 , we recall the notion of Beck-Chevalley cubes and make these as explicit as possible. In section 2.4 , we recall the axioms of A_{n} -schobers. 2.1 Cubical diagrams In this section, we recall cubical diagrams in general. We recall the indexing set \{0,1\}^{n} and its interpretation in terms of compositions of n+1 . Definition 2.1 . A composition of a number n\geq 0 is an ordered tuple of positive natural numbers \sigma=(n_{1},\ldots,n_{k}) summing up to n . In case n=0 the empty tuple serves as the unique composition of n . There are precisely 2^{n} compositions of n+1 . They can be enumerated explicitly in terms of their binary presentation which we recall below. Definition 2.2 . The binary presentation map \psi_{n}:\operatorname{Comp}(n+1)\to\{0,1\}^{n} is defined by \psi_{n}(n_{1},\ldots,n_{k})=0^{n_{1}-1}10^{n_{2}-1}1\ldots 0^{n_{k}-1}. The symbol \psi denotes any such \psi_{n} generically. We have for instance \psi(3,5)=0010000 and \psi(3,5,5,3)=001000010000100 . We remark for later reference that the palindromic nature of a composition \sigma is reflected in its binary presentation \psi(\sigma) . For two compositions \sigma,\tau\in\operatorname{Comp}(n+1) we write \sigma\leq\tau if \tau is a refinement of \sigma . With respect to the lexical partial ordering on \{0,1\}^{n} , the map \psi is an isomorphism. 2.2 Bifactorization cubes In this section, we recall the construction of bifactorization cubes. We make the construction as explicit as possible. As indexing set for the vertices we typically use the binary presentation set \{0,1\}^{n} . The definition of bifactorization cubes starts from a pair of two-element compositions (a,b),(c,d)\in\operatorname{Comp}(n+1) . We start by introducing a case distinction by relation between the numbers a,b,c,d . Lemma 2.3 . Let ((a,b),(c,d))\in\operatorname{Comp}(n+1) be a pair of compositions. Then it is of one of the following forms: • a=c cases 1. ((a,b),(c,d))=((c,c+m),(c,c+m)) with c,m>0 . 2. ((a,b),(c,d))=((a,a),(a,a)) with a>0 . 3. ((a,b),(c,d))=((b+m,b),(b+m,b)) with b,m>0 . • a>c cases 1. ((a,b),(c,d))=((c+l,c),(c,c+l)) with c,l>0 . 2. ((a,b),(c,d))=((b+m+l,b),(b+m,b+l)) with b,m,l>0 . 3. ((a,b),(c,d))=((c+l,c+m),(c,c+m+l)) with c,m,l>0 . • a<c cases 1. ((a,b),(c,d))=((a,a+m),(a+m,a)) with a,m>0 . 2. ((a,b),(c,d))=((a+l,a+m),(a+m+l,a)) with a,d,m>0 . 3. ((a,b),(c,d))=((a,a+m+l),(a+m,a+l)) with a,d,m>0

Read the complete text