Prepotential, mirror map and \(F\)-theory on \(K3\) (Q1298396)

The authors compute certain one-loop corrections to \(F^4\) couplings of \(2n+2\) dimensional \(N=1\) supersymmetric heterotic strings compactified on \(T^2\), generically written as \[ \left. \text{Re}[\tau_{\text{eff}}(T,U)]F\wedge^nF\sim \int \frac{d^2\tau}{\tau_2}Z_{(2,2)}(T,U,q,\bar q) \hat{\mathcal A}_{2n+2}(F,q)\right|_{\text{2n-form}}, \] where \(q=e^{2\pi i\tau}\), \(\bar q=e^{-2\pi i\bar\tau}\), \(T\), \(U\) are the Kähler and complex structure moduli and \(Z_{(2,2)}\) is the partition function of the two-torus \(T^2\). The couplings in the \(T,U\) sub-sector are shown, after evaluation of some integrals, to be derived from an underlying prepotential \(\mathcal G\), \[ \left. (\tau_{\text{eff}})_{ijk\ell}(T,U) \right|_{\text{harmonic part}}= \partial_i\partial_j\partial_k\partial_\ell\mathcal G(T,U), \] with \(i,j,\dots\in\{T,U\}\). Particularly there are discussed the examples with \(E_8\times E_8\) and \(SO(4)^4\) gauge symmetry. Referring to the duality of heterotic compactifications on \(T^2\) and \(F\)-theory on \(K3\) the authors argue that the couplings can be obtained in \(F\)-theory, more precisely from 7-brane geometry in type IIB language. Again, the gauge symmetries \(E_8\times E_8\) and \(SO(4)^4\) serve as an example. For these theories, the relevant geometry can be mapped to Riemann surfaces. Mathematically, this means to associate holomorphic 5-point couplings of the form \(\partial_t^{\;5}\mathcal G\sim\sum g_\ell\ell^5\frac{q^\ell}{1-q^\ell}\) to \(K3\) surfaces. The erratum repeats the computation of the \(\operatorname {Tr}F_\alpha^2 \operatorname {Tr}F_\beta^2\) couplings (for \(\alpha\neq \beta\)), given in eqs. (2.23) and (2.24) of section 2.2.