--- title: Satake Framed Double Quivers url: https://www.emergentmind.com/topics/satake-framed-double-quivers type: topic --- # Satake Framed Double Quivers Searching arXiv for the cited papers and closely related background needed for a precise, research-grounded article. First, I’ll retrieve the main 2017 paper on derived Satake and Coulomb branches. Then I’ll retrieve the 2025 paper introducing Satake framed double quivers and islices. I’ll also check the first companion paper and the Sicilian-theory background explicitly referenced in the source material. Satake framed double quivers are framed double quivers equipped with quasi-split Satake-diagram data, namely an involution $\tau$ on an ADE Dynkin diagram together with a bipartite splitting of the $\tau$-fixed vertices, so that the usual type-$A$ representation spaces are replaced at fixed vertices by orthogonal or symplectic pieces and, in a special quasi-split rank-one situation, by additional $\wedge^2$-blocks. In the 2025 formulation, these data produce symplectic representation spaces ${}^{\imath}E_{V,W}$, gauge and flavor groups $G^\imath(V^\imath)$ and $G^\imath(W^\imath)$, and BFN-type iCoulomb branches $\mathcal M_C^\imath(V^\imath,W^\imath)$ conjectured to normalize top-dimensional components of affine Grassmannian $\sigma$-fixed slices, or islices, ${}^{\imath}\overline{W}_\mu^\lambda$ [2510.10652]. Their background lies in the earlier derived-Satake realization of Coulomb-branch algebras for framed and double quivers through the morphism $T\colon R\to \mathrm{Gr}_G$, where $T_*\omega_R[-2\dim N]$ becomes a commutative ring object in the equivariant derived Satake category and its Ext-algebra recovers the Coulomb branch [1706.02112]. ## 1. Quiver-theoretic definition The starting point is a simple ADE Lie algebra $\mathfrak g$ with Dynkin diagram $I$ and Cartan matrix $C=(c_{ij})$, together with a quasi-split Satake diagram $(I,\tau)$ where $\tau$ is an involutive automorphism of the diagram, $\tau^2=\mathrm{Id}$. The split case is $\tau=\mathrm{Id}$; the genuinely quasi-split cases with $\tau\neq\mathrm{Id}$ are listed as types AIII$_{2r-1}$, AIII$_{2r}$, DI$_r$, and EII$_6$ [2510.10652]. The vertex set is decomposed into $\tau$-orbits by \[ I_0=\{i\in I\mid \tau i=i\},\qquad I_1\ \text{a set of representatives of length-2 orbits},\qquad I_{-1}=\tau(I_1), \] and one sets $I^\imath=I_1\cup I_0$. Starting from an ADE quiver $Q=(I,\Omega)$, one forms its double quiver $\overline Q$ by adjoining a reversed arrow for each $h\in\Omega$, then the framed quiver $Q^f$ by adding framing nodes $I'=\{i'\mid i\in I\}$ and framing arrows $i'\to i$, and finally the framed double quiver \[ \widehat Q=(I\cup I',\,\Omega\cup \overline\Omega\cup \Omega^{\mathrm{fr}}\cup \overline\Omega^{\mathrm{fr}}). \] A Satake framed double quiver is precisely such a framed double quiver together with the quasi-split Satake diagram and the induced involution $\widehat\tau$ on arrows [2510.10652]. The induced involution on arrows is specified by \[ (\widehat\tau h)'=\tau(h'),\qquad (\widehat\tau h)''=\tau(h''), \] with two special rules: if $\tau(h')=h'$ and $\tau(h'')=h''$, then $\widehat\tau h=\overline h$; if $\tau(h')=h''$, then $\widehat\tau h=h$. For organizing the $\tau$-symmetric modifications, the distinguished subsets \[ \Omega_0=\{h\in\Omega\mid h',h''\in I_0\},\qquad \Omega_{1\text{-}0}=\{h\in\Omega\mid h'\in I_1,\ h''\in I_0\}, \] \[ \Omega_{qs}=\{h\in\Omega\mid \tau(h')=h''\}, \qquad I_1^{qs}=\{i\in I_1\mid i\ \text{and}\ \tau i\ \text{are joined by an edge }h\in\Omega\} \] are introduced, with a bijection $I_1^{qs}\to \Omega_{qs}$ sending $i$ to the edge connecting $i$ with $\tau i$ [2510.10652]. A central feature is that the Satake structure modifies not only the combinatorics of the quiver but also the permitted dimension data. The imposed conditions are \[ \dim V_i=\dim V_{\tau i},\qquad \dim W_i=\dim W_{\tau i}\qquad (i\in I_1\cup I_{-1}), \] together with the “not-2-odd” parity conditions \[ \text{either }\dim V_i\text{ or }\dim W_i\text{ is even}\qquad (i\in I_0), \] \[ \text{either }\dim V_{h'}\text{ or }\dim V_{h''}\text{ is even}\qquad (h\in \Omega_0). \] The paper states that this rephrases the parity condition $c_{ij}\theta_i\theta_j=0$ from the iGKLO construction [2510.10652]. ## 2. Orthogonal, symplectic, and hybrid representation data The fixed-point locus $I_0$ is further decomposed into a bipartite partition \[ I_0=I_0^{\oplus}\cup I_0^{\ominus}, \] with the convention that each arrow connects a vertex in $I_0^{\oplus}$ to one in $I_0^{\ominus}$. This determines signs \[ \epsilon_i= \begin{cases} +,& i\in I_0^{\oplus},\\ -,& i\in I_0^{\ominus}, \end{cases} \qquad \overline\epsilon_i=-\epsilon_i. \] These signs control how vector spaces at fixed vertices are replaced by orthogonal or symplectic variants [2510.10652]. For $i\in I_0$, the Satake-modified gauge and framing spaces are \[ V_i^\imath:=V_i^{\epsilon_i}= \begin{cases} V_i^+\text{ (orthogonal)},& i\in I_0^{\oplus},\\ V_i^-\text{ (symplectic, of dimension }2\lfloor \tfrac12\dim V_i\rfloor),& i\in I_0^{\ominus}, \end{cases} \] and \[ W_i^\imath:=W_i^{\overline\epsilon_i}= \begin{cases} W_i^-,& i\in I_0^{\oplus},\\ W_i^+,& i\in I_0^{\ominus}. \end{cases} \] The resulting symmetry groups are \[ G^\imath(V^\imath):=\prod_{i\in I_1}\mathrm{GL}(V_i)\times \prod_{i\in I_0}G^{\epsilon_i}(V_i^\imath), \qquad G^\imath(W^\imath):=\prod_{i\in I_1}\mathrm{GL}(W_i)\times \prod_{i\in I_0}G^{\overline\epsilon_i}(W_i^\imath), \] where $G^{\epsilon_i}(V_i^\imath)$ means $\mathrm{SO}$ for $\epsilon_i=+$ and $\mathrm{Sp}$ for $\epsilon_i=-$ [2510.10652]. The standard framed-double-quiver cotangent representation is \[ E_{V,W} = \bigoplus_{h\in \Omega\cup \overline\Omega}\mathrm{Hom}(V_{h'},V_{h''}) \ \oplus\ \bigoplus_{i\in I}\big(\mathrm{Hom}(W_i,V_i)\oplus \mathrm{Hom}(V_i,W_i)\big), \] but the Satake-ified symplectic vector space is \[ \begin{aligned} {}^\imath E_{V,W} =&\ E_{V^A,W^A}\\ &\oplus\left(\bigoplus_{h\in\Omega_{1\text{-}0}}\mathrm{Hom}(V_{h'},V_{h''}^\imath)\ \oplus\ \bigoplus_{h\in \overline\Omega_{1\text{-}0}}\mathrm{Hom}(V_{h'}^\imath,V_{h''})\right)\\ &\oplus\left(\bigoplus_{h\in\Omega_0}\mathrm{Hom}(V_{h'}^\imath,V_{h''}^\imath)\ \oplus\ \bigoplus_{i\in I_0}\mathrm{Hom}(W_i^\imath,V_i^\imath)\right)\\ &\oplus \bigoplus_{i\in I_1^{qs}}\left(\wedge^2V_i\oplus \wedge^2V_i^*\right). \end{aligned} \] The final wedge term appears precisely in quasi-split type AIII$_{2n}$ when $I_1^{qs}\neq \varnothing$ [2510.10652]. Lemma 8.7 states that ${}^\imath E_{V,W}$ is naturally symplectic with actions of $G^\imath(V^\imath)$ and $G^\imath(W^\imath)$. This is the structural reason that Satake framed double quivers can be inserted into the BFN Coulomb-branch formalism despite generally not being of ordinary cotangent type [2510.10652]. The terminology “orthogonal,” “symplectic,” and “hybrid” is literal rather than metaphorical. In split type, $\tau=\mathrm{Id}$ and the modification keeps the ADE diagram but replaces the fixed-vertex type-$A$ pieces by orthogonal or symplectic ones. In nonsplit type, one obtains type-$A$ data on $I_1$ together with classical $\mathrm{SO}/\mathrm{Sp}$ pieces on $I_0$ and, in AIII$_{2n}$, the extra $\wedge^2$ block. The paper explicitly describes these as “hybrid” representations [2510.10652]. ## 3. iCoulomb branches and affine Grassmannian islices Given the symplectic representation ${}^\imath E_{V,W}$ and gauge group $G^\imath(V^\imath)$, one defines the iCoulomb branch \[ \mathcal M_C^\imath(V^\imath,W^\imath) \] using the Braverman–Finkelberg–Nakajima framework in its general version, namely as a convolution algebra arising from equivariant Borel–Moore homology over the affine Grassmannian of $G^\imath(V^\imath)$ with matter ${}^\imath E_{V,W}$ [2510.10652]. The role played by explicit complex moment map equations $\mu_{\mathbb C}$ in cotangent quivers is here replaced by the BFN convolution definition; the paper emphasizes that the branch is not of cotangent type in general, although in the special AIII$_{2n}$ case it remains of cotangent type with wedge representation. The geometric target of the construction is a $\sigma$-fixed locus inside affine Grassmannian slices. For even spherical $\mu$, the Poisson involution $\sigma$ preserves $W_\mu$, and the fixed point locus \[ {}^\imath W_\mu:=(W_\mu)^\sigma \] is Poisson. If $\tau\lambda=\lambda$, then the slice $\overline W_\mu^\lambda$ is preserved by $\sigma$, and the associated islice is \[ {}^\imath\overline W_\mu^\lambda:=(\overline W_\mu^\lambda)^\sigma. \] The paper states that the islice inherits a Poisson structure via Dirac reduction [2510.10652]. The principal geometric conjecture is Conjecture 8.13(1): the iCoulomb branch $\mathcal M_C^\imath(V^\imath,W^\imath)$ is a normalization of a top-dimensional component of the affine Grassmannian islice ${}^\imath\overline W_\mu^\lambda$ [2510.10652]. The associated dimension numerology is explicit. The open fixed-point piece ${}^\imath U_\mu^\lambda$ is nonempty if and only if the parity condition \[ c_{ij}\theta_i\theta_j=0,\qquad i\neq j\in I, \] holds, and then \[ \dim\,{}^\imath U_\mu^\lambda = 2\sum_{i\in {}^\imath} v_i. \] The same section states that the rank vectors of the component groups of $G^\imath(V^\imath)$ and $G^\imath(W^\imath)$ match the dimension vectors extracted from the coweights $\lambda,\mu$ [2510.10652]. This geometry is best understood as a fixed-point analogue of the ordinary affine Grassmannian-slice/Coulomb-branch correspondence. A plausible implication is that Satake framed double quivers supply the quiver-theoretic model for the $\sigma$-fixed Poisson geometry in much the same way that ordinary framed double quivers model type-$A$ Coulomb branches. The paper makes this suggestion precise only at the level of normalization and top-dimensional components, not as a complete identification in all cases [2510.10652]. ## 4. Derived-Satake antecedents and the type-$A$ prototype The 2017 companion paper provides the derived-Satake mechanism that underlies the later quiver–Satake synthesis. For a complex reductive group $G$ and a finite-dimensional complex representation $N$, the variety of triples $R$ is defined as the moduli of $(P,\sigma,s)$ where $P$ is a trivializable $G$-bundle on the formal disk $D=\mathrm{Spec}\,\mathbb C[[z]]$, $\sigma$ is a trivialization over the punctured disk $D^\times=\mathrm{Spec}\,\mathbb C((z))$, and $s$ is a compatible section of the associated $N$-bundle. Equivalently, points are pairs $(gG_O,s)$ with $g\in G((z))$ and $s\in N[[z]]$ such that both $s$ and $g^{-1}\!\cdot s$ are regular, modulo $G_O=G[[z]]$ [1706.02112]. Projection to the affine Grassmannian defines a $G_O$-equivariant morphism \[ T\colon R\to \mathrm{Gr}_G,\qquad (P,\sigma,s)\mapsto (P,\sigma), \] or, in loop coordinates, $T(gG_O,s)=gG_O$. Writing $\omega_R$ for the dualizing complex on $R$, one sets \[ \mathcal A:=T_*\omega_R[-2\dim N]\in DG(\mathrm{Gr}_G). \] The affine Grassmannian carries the usual convolution product \[ F*G:=m_!(p_1^*F\otimes^L p_2^*G), \] and Proposition 2.1 gives $\mathcal A$ a canonical multiplication $\mathcal A*\mathcal A\to \mathcal A$ and unit $1_{\mathrm{Gr}_G}\to \mathcal A$, making $\mathcal A$ a unital associative algebra object. Theorem 2.5 proves commutativity, using nearby cycles on the Beilinson–Drinfeld Grassmannian; Appendix B gives a direct global proof via a global convolution diagram [1706.02112]. The associated Coulomb-branch algebra is recovered as Ext-cohomology: \[ \operatorname{Ext}^*_{DG(\mathrm{Gr}_G)}(1_{\mathrm{Gr}_G},\mathcal A)\cong H_*^{G_O}(R), \] and the induced multiplication agrees with the Coulomb-branch convolution product \[ [\alpha]*[\beta]:=m_*\big(p_1^*[\alpha]\cap p_2^*[\beta]\big). \] More generally, if $\mathcal A$ is any commutative ring object in $DG(\mathrm{Gr}_G)$, then \[ \mathcal M_C(\mathcal A):=\operatorname{Spec}\operatorname{Ext}^*(1_{\mathrm{Gr}_G},\mathcal A) \] is a commutative graded Coulomb-branch algebra, and gluing is implemented by diagonal pullback \[ \mathcal A_{\mathrm{glued}}:=\Delta^*(\boxtimes_i \mathcal A_i), \] which again yields a commutative ring object [1706.02112]. This framework is the type-$A$ precursor of the later Satake framed double-quiver picture. In particular, the paper studies star-shaped framed type-$A$ quivers, expected to be Higgs branches of $3d$ Sicilian theories, and shows that gluing and “leg amputation” on ring objects produce their Coulomb branches [1706.02112]. The main identification in this direction is Theorem 2.11: for the framed type-$A$ quiver with $\dim V=(N-1,N-2,\ldots,1)$ and $\dim W=(N,0,\ldots,0)$ with flavor group $\mathrm{PGL}(N)$, \[ T_*\omega_R[-2\dim N]\cong \mathcal A_R \] as ring objects on $\mathrm{Gr}_{\mathrm{PGL}(N)}$, where $\mathcal A_R$ is the regular sheaf corresponding under geometric Satake to the regular representation $\mathbb C[\check G]$. Consequently, \[ \operatorname{Ext}^*(1_{\mathrm{Gr}},\mathcal A_R)\cong H_*^{G_O}(R)\cong \mathbb C[\mathcal N], \] recovering the nilpotent cone of the Langlands dual group [1706.02112]. The Satake framed double-quiver construction of 2025 can be read as an extension of this paradigm from ordinary framed/double quivers to quasi-split and fixed-point data. That is an interpretation rather than an explicit theorem, but it matches the stated “bridging” role of the 2017 paper: from framed double quivers one obtains ring objects on the affine Grassmannian, and from ring objects one obtains Coulomb branches through Ext-cohomology [1706.02112]. ## 5. Shifted twisted Yangians and quantization On the algebraic side, Satake framed double quivers are linked to shifted twisted Yangians. For any quasi-split Satake diagram $(I,\tau)$ of ADE type and any even spherical coweight $\mu$, the shifted twisted Yangian ${}^\imath Y_\mu$ is defined as the $\mathbb C$-algebra generated by Drinfeld currents \[ H_i^{(r)},\ B_i^{(s)}\qquad (i\in I,\ r\in\mathbb Z,\ s\in \mathbb Z_{>0}), \] subject to the relations recorded in Definition 2.2 and equations (2.7)–(2.10), including the initial conditions \[ H_i^{(r)}=0\ \text{for }r<-\langle \mu,\alpha_i\rangle,\qquad H_i^{(-\langle \mu,\alpha_i\rangle)}=1 \] and the commutativity $[H_i^{(r_1)},H_j^{(r_2)}]=0$ [2510.10652]. Theorem 2.10 establishes a PBW basis: ordered monomials in root vectors $B_\beta^{(r)}$ together with Cartan coefficients $H_i^{(r)}$ form a PBW basis of ${}^\imath Y_\mu$. The same theorem gives injective shift homomorphisms \[ \iota^\tau_{\mu,\nu}\colon {}^\imath Y_\mu \to {}^\imath Y_{\mu+\nu+\tau\nu} \] for anti-dominant $\nu$ [2510.10652]. The iGKLO representation is the mechanism relating these algebras to geometry. Theorem 3.8 states that for a quasi-split Satake diagram, a $\tau$-invariant dominant coweight $\lambda$, and even spherical $\mu$ satisfying the parity constraint $c_{ij}\theta_i\theta_j=0$, there is a homomorphism \[ \Phi_\mu^\lambda\colon {}^\imath Y_\mu[\mathbf z]\longrightarrow \mathcal A, \] where $\mathcal A$ is a ring of difference operators generated by $w_{i,r}$, shifts $S_{i,r}^{\pm1}$, and minors $(w_{i,r}\pm w_{i,r'}+m)^{-1}$ with relations \[ [S_{i,r}^{\pm1},w_{j,r'}]=\pm\delta_{ij}\delta_{r,r'}S_{i,r}^{\pm1},\qquad [w_{i,r},w_{j,r'}]=[S_{i,r},S_{j,r'}]=0. \] The truncated shifted twisted Yangian is defined by \[ {}^\imath Y_\mu^\lambda:=\operatorname{Im}\Phi_\mu^\lambda \] [2510.10652]. The fixed-point geometry reappears in the classical limit. Theorem 7.4 identifies the associated graded of ${}^\imath Y_\mu$ with the coordinate ring of the Poisson fixed locus: \[ \mathrm{gr}^{F_{\mu_1}^\bullet}({}^\imath Y_\mu)\cong \mathbb C[{}^\imath W_\mu]. \] Theorems 7.5 and 7.6 further state that the classical limit of the iGKLO map defines the closure $\overline C_\mu^\lambda$ of a top-dimensional irreducible component $C_\mu^\lambda\subseteq {}^\imath U_\mu^\lambda$, both in the undeformed case and over the Beilinson–Drinfeld base [2510.10652]. The paper therefore proposes the quantization conjecture that ${}^\imath Y_\mu^\lambda$ quantizes $\overline C_\mu^\lambda$. This algebraic picture explains why Satake framed double quivers are not merely a geometric rephrasing of fixed-point slices. They provide the quiver input for iCoulomb branches, while shifted twisted Yangians provide a quantization expected to match the same geometric objects. The 2025 paper formulates this as a program rather than a completed equivalence in full generality [2510.10652]. ## 6. Type AI, explicit examples, and current status The best-developed identifications occur in type AI. In that case, with $\tau=\mathrm{Id}$ and $\mu$ dominant and even, a variant of ${}^\imath Y_\mu^{N\varpi_1^\vee}$ is identified with the truncated shifted twisted Yangian $Y_{n,\ell}^+(\sigma)$, and hence with finite $W$-algebras of classical types BCD [2510.10652]. Geometrically, Theorem 7.9 identifies the islices with nilpotent Slodowy-slice intersections: \[ {}^\imath W_\mu^\lambda \cong \mathbb O_{\pi_1}^\epsilon\cap \mathcal S_{\pi_2}^\epsilon,\qquad {}^\imath\overline W_\mu^\lambda \cong \overline{\mathbb O_{\pi_1}^\epsilon\cap \mathcal S_{\pi_2}^\epsilon}, \] where $\epsilon=+$ or $-$ according to the parity of parts of $\pi_2$. The statement generalizes Lusztig’s bijection and the Mirković–Vybornov isomorphisms from type $A$ to classical types B/C/D through the $\sigma$-fixed-locus formalism [2510.10652]. Several explicit low-rank examples clarify the construction. For type $A_1$ with $G=\mathrm{PGL}_2$ and $\tau=\mathrm{Id}$, the slice $\overline W_\mu^\lambda$ is described by polynomial matrices \[ \begin{pmatrix} d(z) & b(z)\\ c(z) & a(z)\end{pmatrix} \] with $a,b,c,d\in\mathbb C[z]$, $a$ monic of degree $v$, $\deg b,\deg c