Hamiltonian reduction for affine Grassmannian slices and truncated shifted Yangians (Q2118924)

Generalized affine Grassmannian slices provide geometric realizations for weight spaces of representations of semisimple Lie algebras. They are also Coulomb branches, symplectic dual to Nakajima quiver varieties. That is, let \(G\) be a complex semisimple group and consider the affine Grassmannian \(\mathrm{Gr}_G = G((t^{-1}))/G[t]\) of \(G\). There are the spherical Schubert cells \(\mathrm{Gr}^\lambda := G[t]t^\lambda\) in the affine Grassmannian for a dominant coweight \(\lambda\) of \(G\). Let \(\mathcal{W}_\mu := G_1[[t^{-1}]]t^\mu\) be transverse orbits for a dominant coweight \(\mu\). One can then construct and study affine Grassmannian slices, which are their intersections \(\overline{\mathcal{W}}^\lambda_\mu := \mathcal{W}_\mu \cap \overline{\mathrm{Gr}^\lambda}\). \textit{A. Braverman} et al. [Adv. Theor. Math. Phys. 23, No. 1, 75--166 (2019; Zbl 1479.81044)] generalized these spaces \(\overline{\mathcal{W}}_\mu^\lambda\) and \(\mathcal{W}_\mu\) to the case where \(\mu\) is not dominant. When \(G\) is simply laced, then \(\overline{\mathcal{W}}^\lambda_\mu\) is the Coulomb branch of a quiver gauge theory, whose Higgs branch is a Nakajima quiver variety. This was extended to the non-simply-laced case, to describe \(\overline{\mathcal{W}}^\lambda_\mu\) as the Coulomb branch of a quiver gauge theory with symmetrizers. The authors prove that neighboring generalized affine Grassmannian slices are related by Hamiltonian reduction by the action of the additive group (Theorem 1.1, page 2): letting \(\lambda\) to be a dominant coweight, \(\mu\) a coweight, and \(\alpha_i^\vee\) any simple coroot, there is a Poisson isomorphism \(\overline{\mathcal{W}}^\lambda_\mu /\!\!/_1 \mathbb{G}_a \cong \overline{\mathcal{W}}^\lambda_{\mu + \alpha_i^\vee}\). This results in Theorem 1.3 (page 3) for shifted Yangians \(Y_{\mu}\), which are used to quantize the spaces \(\mathcal{W}_{\mu}\): upon applying associated graded, the comultiplication map, \( Y_{\mu_1 + \mu_2} \rightarrow Y_{\mu_1} \otimes Y_{\mu_2}\) becomes the multiplication map \(\mathbb{C}[\mathcal{W}_{\mu_1 + \mu_2}] \rightarrow \mathbb{C}[\mathcal{W}_{\mu_1}] \otimes \mathbb{C}[\mathcal{W}_{\mu_2}]\). They also prove Theorem 1.4 (page 3): comultiplication gives an isomorphism \(Y_\mu[(E_i^{(1)})^{-1}] \cong Y_{-\alpha_i^\vee}^0 \otimes Y_{\mu + \alpha_i^\vee}\), obtaining \(Y_{\mu + \alpha_i^\vee}\) as the quantum Hamiltonian reduction of \(Y_\mu\). Here, \(E_i^{(1)}\) is inverted since it is the quantum moment map for the \(\mathbb{G}_a\)-action. The algebra \(Y_{-\alpha_i^\vee}^0\) is isomorphic to the ring of differential operators on \(\mathbb{C}^\times\), a quantum version of the isomorphism \( \overline{\mathcal{W}}^0_{-\alpha_i^\vee} \cong T^* \mathbb{C}^\times\). Finally, the authors also prove a similar version for their quantization algebras known as truncated shifted Yangians for \(G=SL_n\) unconditionally and for general \(G\), with certain conditions on the reducedness conjecture (Theorem 1.5, page 4): comultiplication gives an isomorphism \(Y_\mu^{\lambda}[(E_i^{(1)})^{-1}] \cong Y_{-\alpha_i}^0 \otimes Y_{\mu + \alpha_i^\vee}^{\lambda}\), resulting in \(Y_{\mu + \alpha_i^\vee}^{\lambda}\) as the quantum Hamiltonian reduction of \(Y_\mu^{\lambda}\).