Quantum cohomology and crepant resolutions: a conjecture (Q2391384)

The paper under review provides an elegant expository account of the crepant resolution conjecture relating the quantum cohomology of a Gorenstein orbifold to the quantum cohomology of its crepant resolutions. Givental encodes all genus 0 Gromov-Witten invariants of a Gorenstein orbifold \(\mathcal{X}\) in the germ \(\mathcal{L}_{\mathcal{X}}\) of a Lagrangian submanifold in a symplectic vector space \(\mathcal{H}_{\mathcal{X}}\). Suppose that \(Y\) is a crepant resolution of \(\mathcal{X}\). The crepant resolution conjecture developed by Coates-Iritani-Tseng and Ruan then states that there is a linear symplectic linear isomorphism \(\mathbb{U}:\mathcal{H}_{\mathcal{X}}\to \mathcal{H}_{Y}\) such that after analytic continuation of \(\mathcal{L}_{\mathcal{X}}\) and \(\mathcal{L}_{Y}\) we have \(\mathbb{U}(\mathcal{L}_{\mathcal{X}})=\mathcal{L}_{Y}\). Here, it is shown that the cohomological crepant resolution conjectures as well as Ruan and Bryan-Graber's versions of the crepant resolution conjecture can be deduced from this statement. By studying the quantization \(\widehat{\mathbb{U}}\) of the symplectic transformation \(\mathbb{U}\), the conjecture above is generalized to the higher-genus version \(\widehat{\mathbb{U}}(\mathcal{D}_{\mathcal{X}})=\mathcal{D}_{Y}\), where \(\mathcal{D}_{\mathcal{X}}\) and \(\mathcal{D}_{Y}\) are the full descendant Gromov-Witten potential of \(\mathcal{X}\) respectively \(Y\). Finally, the paper addresses the problem of specializing the exceptional quantum parameters in the crepant resolution conjecture. This is obtained by interpreting the values of the specialization as coming from a shift in the base point on Givental's cone.