Inverse Hamiltonian Reduction for affine W-algebras in type A
Dylan Butson and Sujay Nair · 2025
Read Inverse Hamiltonian Reduction for affine W-algebras in type A by Dylan Butson and Sujay Nair (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-2508-18248.txt.
Introduction
Dylan Butson and Sujay Nair give a geometric proof of inverse Hamiltonian reduction for affine W-algebras in type A, in a 2025 paper. These algebras arise from symmetry structures in physics and geometry. The result embeds the algebra tied to a smaller nilpotent orbit into the one tied to a larger orbit, paired with a free field algebra.
Their route builds strict quantizations of special slices tied to group action. They form sheaves of vertex algebras over the arc spaces of those slices, then localize them on near flat open sets. A broader version covers reduction of general vertex algebra objects in a standard category. The work firms up a key duality in the field.
Catalog record
| CREATOR | Dylan Butson and Sujay Nair |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 88 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-18248
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Inverse Hamiltonian Reduction for affine W-algebras in type A Dylan Butson and Sujay Nair Abstract . We give a geometric proof of inverse Hamiltonian reduction for all affine W-algebras in type A at generic level, a certain embedding of the affine W-algebra corresponding to an arbitrary nilpotent in 𝔤 𝔩 N \mathfrak{gl}_{N} into that corresponding to a larger nilpotent with respect to the closure order on orbits, tensored with an auxiliary algebra of free fields. We proceed by constructing strict chiral quantizations of equivariant Slodowy slices, sheaves of ℏ \hbar -adic vertex algebras on the arc spaces of the slices, and then localizing them on quasi-Darboux open sets. We also provide a generalization for the Drinfeld–Sokolov reduction of arbitrary vertex algebra objects in the Kazhdan–Lusztig category. 1. Introduction Given a nilpotent element \chi in \mathfrak{g}^{*}=\mathfrak{gl}_{N}^{*} and a level k\in\mathbb{C} , one can construct a certain vertex algebra, {\mathcal{W}}^{k}(\mathfrak{g},\chi) , called the affine W-algebra. In fact, up to isomorphism this vertex algebra only depends on the nilpotent orbit of \chi , and so can be labeled by \mu=[\mu_{1}\geq\mu_{2}\geq\mu_{3}\geq\dots\mu_{n-1}\geq 0] a partition of N , and we identify the set of all such partitions of length n-1 with the set of dominant, integral coweights \Lambda^{+}(\mathrm{PGL}_{n}) of \mathrm{PGL}_{n} . For the special case where \mu=[1^{N}] , i.e. , \chi=0 , this vertex algebra is the universal affine vertex algebra V^{k}(\mathfrak{g}) . One has a homological construction of W-algebras for \mu\geq[1^{N}] by quantum Hamiltonian (or Drinfeld–Sokolov) reduction of V^{k}(\mathfrak{g}) , see [ KRW03 ] . This construction has proven to be of great value in studying the representation theory of these algebras [ Ara07 ] . Recently, for \mu_{1}\geq\mu_{2} satisfying certain hypotheses, it has been shown by [ GJ25 ] that one can obtain {\mathcal{W}}^{k}_{\mu_{1}} from {\mathcal{W}}^{k}_{\mu_{2}} by a kind of partial Hamiltonian reduction. The main result of this work concerns inverse Hamiltonian reduction , which realizes {\mathcal{W}}^{k}_{\mu_{2}} as a subalgebra of the tensor product of {\mathcal{W}}^{k}_{\mu_{1}} with so-called free fields. Theorem 1.1 . Let \mu\in\Lambda^{+}(\mathrm{PGL}_{n}) and \alpha a positive co-root such that \mu+\alpha\in\Lambda^{+}(\mathrm{PGL}_{n}) . Then, at generic level k , we have an embedding of \hbar -adic vertex algebras (1.0.1) {\mathcal{W}}_{\hbar,\mu}^{k}\hookrightarrow{\mathcal{W}}^{k}_{\hbar,\mu+\alpha}\hat{\otimes}{\mathcal{D}}^{ch}_{\rm loc}(\mathbb{A}^{\rm len\,\alpha})~, for {\mathcal{D}}^{ch}_{\rm loc}(\mathbb{A}^{\rm len\,\alpha}) a localization of the \hbar -adic vertex algebra of chiral differential operators on \mathbb{A}^{\rm len\,\alpha} . This theorem completes the generalization to type A of a family of results going back to [ Sem94 ] , as we will explain below. It is obtained in much the same way as in [ BN25a ] , where we established the analogous result for finite W-algebras. Indeed, crucial ingredients in the proof are the geometric results on equivariant Slodowy slices obtained in loc. cit. . The other crucial ingredient, like in finite type, is deformation theory—now of vertex algebras as opposed to associative algebras. In a sense, this work is an application of the results of [ BN25b ] on deformations of quantizations of arc spaces. Our proof proceeds by first constructing quantizations of the arc spaces of the equivariant Slodowy slices. Namely, for each k\in\mathbb{C} , there is a sheaf of \hbar -adic vertex algebras {\mathcal{A}}^{k}_{\hbar,G,\mu} on {\mathcal{J}}S_{G,\mu} whose \hbar -adic vertex algebra of global sections is isomorphic to the \hbar -adic analogue of the equivariant affine W-algebra at level k of [ Ara18 ] , that is, \Gamma({\mathcal{J}}S_{G,\mu},{\mathcal{A}}^{k}_{\hbar,G,\mu})\cong{\bf W}^{k}_{\hbar,G,\mu}~. We then prove inverse Hamiltonian reduction for these equivariant W-algebras, constructing the inclusion by localizing this sheaf on the image of the inverse Hamiltonian reduction chart of [ BN25a ] and appealing to the deformation theory results of [ BN25b ] . Theorem 1.2 . Let \mu\in\Lambda^{+}(\mathrm{PGL}_{n}) and \alpha a positive co-root such that \mu+\alpha\in\Lambda^{+}(\mathrm{PGL}_{n}) . Then, at generic level k , we have an embedding of \hbar -adic vertex algebras (1.0.2) {\bf W}_{\hbar,G,\mu}^{k}\hookrightarrow{\bf W}^{k}_{\hbar,G,\mu+\alpha}\hat{\otimes}{\mathcal{D}}^{ch}_{\rm loc}(\mathbb{A}^{\rm len\,\alpha})~. Inverse hamiltonian reduction for the ordinary affine W-algebras follows by taking G({\mathcal{O}}) -invariants. The result for equivariant W-algebras in fact implies a stronger version of inverse Hamiltonian reduction for arbitrary modules in the Kazhdan–Lusztig category. Namely, we prove the following: Theorem 1.3 . Let M\in{\rm KL}_{k} , and denote by H^{0}_{\rm DS}(M,\mu) the Drinfeld–Sokolov reduction of M at some representative of the orbit labeled by \mu . Then we have natural embeddings of {\mathcal{W}}^{k}_{\hbar,\mu} modules (1.0.3) H^{0}_{\rm DS}(M,\mu)\hookrightarrow H^{0}_{\rm DS}(M,\mu+\alpha)\hat{\otimes}{\mathcal{D}}^{ch}_{\rm loc}(\mathbb{A}^{\rm len\,\alpha})~, where the {\mathcal{W}}_{\hbar,\mu}^{k} -module structure on the right comes from the embedding in Theorem 1.1 . In particular, if M is a vertex algebra object in \mathrm{KL}_{k} , this is an embedding of vertex algebras. 1.1. Relation to other work The first example of inverse Hamiltonian reduction for affine W-algebras was the work of Semikhatov [ Sem94 ] , which considered the \mathfrak{sl}_{2} case. More recently, higher rank and superalgebra examples have been constructed by Adamović in [ Ada05 , Ada19 ] . The first somewhat systematic proof of inverse Hamiltonian reduction in type A is due to Fehily [ Feh23 ] , who proved the statement for hook-type W-algebras using the free-field realizations of these W-algebras from [ Gen17 ] . This approach was extended to all examples in rank three in [ CFLN25 , FFFN24 ] , and further to all Virasoro-type reductions in [ FKN25 ] . Additionally, the present work also offers a novel proof of what is known in the literature as reduction by stages. This is a result first conjectured, for finite type, in Morgan’s thesis [ Mor15 ] and proven in the finite and affine case by [ GJ24 , GJ25 ] . We note, however, that our hypotheses differ from those of loc. cit. in Type A. In a sense, inverse Hamiltonian reduction can be viewed as a special case of a generalized coproduct on truncations of the shifted affine Yangian of \mathfrak{gl}_{n} . For the finite Yangian, this comultiplication was described in [ FKP+18 ] . In the affine case, however, constructing the general coproduct has proven to be a challenging endeavor. In the case of rectangular nilpotents, some results have been obtained in [ KU22 , GRZ23 ] and geometric approaches to this problem can be found in [ BR23 , But23 ] . Unlike [ BN25a ] , we make almost no reference to the affine Grassmannian, or generalized slices therein. While the results of our previous work related various quantizations of certain generalized slices in the affine Grassmannian to each other, this work relates strict chiral quantizations of those same slices. One wonders whether this may have an interpretation in terms of the geometry of the double affine Grassmannian, (see [ BF10 ] ), but we leave this speculation to future work. Finally, this work is yet another entry into the ever-growing list of geometric free-field realizations. The idea of obtaining free-field realizations by constructing a sheaf of vertex algebras and localizing them onto a global (quasi-)Darboux patch originates in [ BMR19 ] . More recently, this viewpoint has been used successfully to provide free-field realizations for several vertex algebras originating from supersymmetric gauge theories—see [ AKM23 , CSYZ25 , BF25 , BN23 , Fur23 ] for an incomplete list. 2. Preliminaries 2.1. Vertex algebras and vertex Poisson algebras Write \mathbb{C}[{{\partial}},z] for the global sections of differentials operators on \mathbb{A}^{1} , and write \mathbb{C}[{{\partial}}] for