All-orders moduli for type II flux backgrounds

Abstract: We investigate the old problem of determining the exact bulk moduli of generic $\mathrm{SU}(3)$-structure flux backgrounds of type II string theory. Using techniques from generalised geometry, we show that the infinitesimal deformations are counted by a spectral sequence in which the vertical maps are either de Rham or Dolbeault differentials (depending on the type of the exceptional complex structure (ECS)) and the horizontal maps are linear maps constructed from the flux and intrinsic torsion. Our calculation is exact, covering all possible supergravity $\mathrm{SU}(3)$-structure flux backgrounds including those which are not conformally Calabi--Yau, and goes beyond the usual linear approximation in three important ways: (i) we allow for finite flux; (ii) we consider perturbative higher-derivative corrections to the supergravity action; and (iii) we consider obstructions arising from higher-order deformations. Despite these extensions we find that the spectral sequence reproduces the na\"ive expectations that come from considering the effective superpotential in the small-flux limit. In particular, by writing the moduli in a form that is independent of the K\"ahler potential on the space of ECSs, and arguing the superpotential does not receive higher-derivative corrections, we show that the spectral sequence is perturbatively exact. Further, preliminary results show that a Tian--Todorov-like lemma implies that all the obstructions vanish. This has implications for the tadpole conjecture, showing that such perturbative, higher-order effects do not provide a way of circumventing the bound.