Quantum immanants and higher Capelli identities (Q1924191): Difference between revisions

Everything is over a field \(k\) of characteristic zero. Let \(E_{ij}\) be the standard generators of \(U(\text{gl}(n))\), let \(C\) be the row determinant \[ \sum_{s\text{ in }S_n} \text{sign}(s) E_{1,s(1)}(E_{2,s(2)}+\delta_{2,s(2)})\dots(E_{n,s(n)}+(n-1) \delta_{n,s(n)}) \] of the matrix \(E+\text{diag}(0,1,\dots,n-1)\), \(E=(E_{ij})\). Represent \(U(\text{gl}(n))\) on \(k[M(n)]\), \(M(n)\) the \(n \times n\) matrices, by \(L(E_{ij})=\sum_a x_{ia} \partial_{ja}\), i.e., \(L(E) = XD'\), where \(X=(X_{ij})\), \(D=(\partial_{ij})\). The Capelli identity says that \(L(C)=(\det X)(\det D)\). \(C\) is a central element of \(U(\text{gl}(n))\). In this paper, the author generalizes \(C\) to what he calls quantum immanants, replacing \(C\) and \(L(C)\) by the trace of an arbitrary polynomial representation of GL\((n)\). These are used to prove higher Capelli identities. His approach is based on \(R\)-matrix formalism.