Koszul duality between Higgs and Coulomb categories $\mathcal{O}$

TL;DR

This paper establishes a Koszul duality between categories related to Higgs and Coulomb branches, linking their algebraic structures and confirming aspects of the symplectic duality conjecture.

Contribution

It proves a Koszul duality theorem connecting Coulomb and Higgs categories of $ ext{O}$, relating them via explicit combinatorial models and confirming key symplectic duality conjectures.

Findings

Established a functor as an equivalence in special cases like Nakajima quiver varieties.

Connected Coulomb and Higgs categories through explicit combinatorial presentations.

Confirmed the symplectic duality conjecture components for these categories.

Abstract

We prove a Koszul duality theorem between the category of weight modules over the quantized Coulomb branch (as defined by Braverman, Finkelberg and Nakajima) attached to a group $G$ and representation $V$ and a category of $G$-equivariant D-modules on the vector space $V$. This is proven by relating both categories to an explicit, combinatorially presented category. These categories are related to generalized categories $\mathcal{O}$ for symplectic singularities. Letting $\mathcal{O}_{\operatorname{Coulomb}}$ and $\mathcal{O}_{\operatorname{Higgs}}$ be these categories for the Coulomb and Higgs branches associated to $V$ and $G$, we obtain a functor $\mathcal{O}_{\operatorname{Coulomb}}^!\to \mathcal{O}_{\operatorname{Higgs}}$ from the Koszul dual of one to the other. This functor is an equivalence in the special cases where the hyperk\"ahler quotient of $T^*V$ by $G$ is a Nakajima quiver variety or smooth hypertoric variety. This includes as special cases the parabolic-singular Koszul duality of category $\mathcal{O}$ in type A, the categorified rank-level duality proposed by Chuang and Miyachi and proven by Shan, Vasserot and Varagnolo, and the hypertoric Koszul duality proven by Braden, Licata, Proudfoot and the author. We also show that this equivalence intertwines so-called twisting and shuffling functors. This together with the duality discussed confirms the most important components of the symplectic duality conjecture of Braden, Licata, Proudfoot and the author in this case.