Uniqueness theorem of \(\mathcal W\)-constraints for simple singularities (Q381010): Difference between revisions

Gromov-Witten invariants satisfy complicated recurrance relations that form a very powerful computational tool. The Gromov-Witten invariants of a point capture interesting information about the cohomology of the moduli space of stable curves. They can be encoded into the partition function of a two-dimensional gravity. This is a formal power series in the variables \(t^k\) for non-negative integer \(k\), and it is completely determined by an infinite collection of linear differential constraints known as the Virasoro constraints. This model can be viewed as a Landau-Ginzberg model for the \(A_1\) singularity. It is natural to generalize to other singularities. Bakalov and Milanov proved that the generating function for analagous invariants for ADE singularities satisfies an infinite collection of linear differential relations, known as the \(\mathcal{W}\)-constraints. This paper begins by reviewing these constraints. It continues to prove that the solution to these relations is uniquely determined up to a constant multiple. The authors then explain how this allows one to recursively compute all of the associated invariants. As an application, they compute the invariants associated to the \(D_4\) singularity.