Quantum hyperbolic geometry (Q2464776)

Kashaev's Hypebolic Volume Conjecture inspired a lot of research on the connection between hyperbolic geometry and quantum topology. One way to approach the issue are the quantum hyperbolic invariants of \textit{S. Baseilhac} and \textit{R. Benedetti} [Geom. Topol. 9, 493--569 (2005; Zbl 1093.57005)]. Ever since they were defined the natural question was whether they arise from a Topological Quantum Field Theory (TQFT) like the older Turaev-Viro invariants. This foundational paper delivers a positive answer to this question with a bonus. The authors construct a family of Quantum Hyperbolic Field Theories (QHFT) that produce a new class of invariants, the quantum hyperbolic ones being a particular case. The QHFT functor \(\mathcal{H}_N\) depends on a choice of a positive integer \(N\) and acts on triples \((Y,L,\rho)\). Here \(Y\) is a compact oriented \(3\)-manifold with boundary, \(L\) is an embedded framed tangle that intersects every boundary component, and \(\rho\) is a representation of \(\pi_1(Y\backslash L)\) in \(PSL_2\mathbb{C}\). Each boundary component has a vector space assigned to it and the amplitude \(\mathcal{H}_N(Y,L,\rho)\) is an element in the tensor product of these spaces defined up to sign and multiplication by \(N\)-th roots of unity. When \(Y\) is closed one obtains numerical invariants of manifolds with links. When \(L\) is a hyperbolic link one can think of them as being invariants of cusped manifolds \(Y\backslash L\) colored by \(\rho\). If \(L\) is unframed and meridian loops are mapped into identity by \(\rho\) one recovers the quantum hyperbolic invariants. When \(N=1\) the link is irrelevant and one recovers the Cheeger-Chern-Simons invariants of \((Y,\rho)\). In fact the authors define a broader class of triangulated \(3\)-pseudomanifolds with `universal quantum hyperbolic functors' and convert \((Y,L,\rho)\) into such a pseudomanifold uniquely up to isomorphism. Construction of the QHFT functor is similar in spirit to the Turaev-Viro construction. Amplitudes are defined by state summing over decorated triangulations. A distinct feature is the essential use of hyperbolic geometry. The triangulation tetrahedra are interpreted as sitting inside the \(3\)-dimensional hypebolic space with vertices on its conformal boundary (complex projective line). The hyperbolic structure is encoded by decorating the edges with cross-ratios. Instead of quantum \(6j\) symbols of Turaev-Viro the authors use as building blocks \(N\times N\) matrix dilogarithms that generalize the scalar one of Rogers. They are related to Kashaev's \(6j\) symbols and can be interpreted as transition functions between sections of line bundles over the moduli of cross-ratios. Dilogarithms have branch cuts and additional decorations (charge and flattening of \textit{W. D. Neumann} [Geom. Topol. 8, 413--474 (2004; Zbl 1053.57010)]) are required to define them unambiguously. As an example the authors compute \(\mathcal{H}_N(S^3,K,\rho)\), where \(K\) is the figure eight knot and \(\rho\) are injective. An application to constructing new moderately projective (defined up to root of unity multiples) representations of mapping class groups is given. More involved applications to the Hyperbolic Volume Conjecture and Turaev's homotopic QFT are left to future work.