In superstring theory, one usually considers compactifications on Calabi-Yau 3-manifolds. These manifolds are in particular compact Kähler, hence possess a Kähler class which gives rise to nontrivial cohomology classes in every even degree. To see this, note that the Kähler class $\omega$ on $M$ is closed by definition, hence if $\omega^k=d\alpha$ for some $(2k-1)$-form $\alpha$, we find that $\omega^{k+1}=\omega\wedge d\alpha=d(\omega\wedge \alpha)$, therefore $\omega^{k+1}$ would be exact as well. But $\omega^n$, where $n=\operatorname{dim}_{\Bbb C}M$, is a volume form, hence by Stokes' theorem it cannot be exact. Thus, $\omega^k$ is not exact for any $k\leq n$. Equivalent, we have the following condition on the Hodge numbers: $h^{p,p}(X)\geq 1$. Now, for my actual question:
Since the powers of the Kähler class always generate nontrivial cohomology classes, these can in some sense be called universal. I was wondering if there is a nice interpretation of these classes in string theory.
I vaguely recall that one can interpret the cohomology classes of the Calabi-Yau manifold that one compactifies on in terms of the multiplets (under supersymmetry) of the resulting effective four-dimensional description, and in particular I'm hoping that the Kähler class gives rise to some kind of universal multiplets.
