\(\widehat{Z}\) and splice diagrams (Q6924359)

The results of the paper fall within the broader program of relating low-dimensional topology to singularity theory, a connection that has proved highly fruitful in the past. This work constitutes a meaningful and valuable contribution to that program as follows.\N\NConsider three-dimensional manifolds \(Y\) that arise as links of normal surface singularities. To such \(Y\)'s one classically assigns a combinatorial object -- the splice diagram -- which captures substantial geometric and topological information about both the singularity and its link.\N\NAn important invariant of 3-manifolds is the \(q\)-series invariant \(\hat{Z}_\sigma(Y;q)\), introduced by Gukov-Pei-Putrov-Vafa [\textit{S. Gukov} et al., J. Knot Theory Ramifications 29, No. 2, Article ID 2040003, 85 p. (2020; Zbl 1448.57020)]. This invariant is associated to a quantum group and a choice of \(\text{spin}^c\) structure \(\sigma\) on \(Y\). In the present paper, the authors fix the gauge group \(G = SU(2)\) and the corresponding quantum group.\N\NThe first main result shows that the series \(Z_0 = \sum_\sigma \widehat{Z}_\sigma\) depends only on the splice diagram of \(Y\), with only a mild additional dependence on the Casson-Walker invariant. As a consequence, if two manifolds have the same universal cover -- a phenomenon that occurs for several important classes of manifolds -- then their \(Z_0\)-invariants (essentially) coincide.\N\NThe second main theorem provides explicit formulas for \(Z_0\) and for the anti-canonical invariants \(\hat{Z}_{{anti{-}canonical}}\) in the case of Seifert manifolds.\N\NIn addition, the authors investigate moduli spaces of \(SL_2(\mathbb C)\) connections on Seifert manifolds and explore their relationship with the spectra of surface singularities.