Twistor spaces of K3 surfaces are non-Kähler compact complex manifolds which play a fundamental role in the moduli theory of K3 surfaces. They come equipped with a holomorphic submersion to the complex projective line which under the period map corresponds to a twistor line in the K3-period domain. In this talk I will explain how one can view a twistor line as a certain base point in the linear cycle space of the period domain. Then, based on joint work in progress with Daniel Greb, Tim Kirschner and Martin Schwald I will present new results concerning the deformations of twistor spaces of K3 surfaces and their relation to the cycle space of the period domain.[-]

I will present a joint work with Sara Angela Filippini, Laurent Manivel and Fabio Tanturri (arXiv: 1704.01436). We introduce a new class of varieties, called orbital degeneracy loci. The idea is to use any orbit closure in a representation of an algebraic group to generalise the classical construction of degeneracy loci of morphisms between vector bundles, and of zero loci as well. After giving the definition of an orbital degeneracy locus, I will explain how to control the canonical bundle of these varieties: under some Gorenstein condition on the orbit closure, it is possible to construct examples of varieties with trivial canonical bundle or of Fano type. Finally, if time will permit, I will give some explicit examples of such degeneracy loci, which allow to construct many Calabi-Yau varieties of dimension three and four, and some new Fano fourfolds.[-]

Hyperkähler manifolds are higher-dimensional analogs of K3 surfaces. Verbitsky and Markmann recently proved that their period map is an open embedding. In a joint work with E. Macri, we explicitly determine the image of this map in some cases. I will explain this result together with a nice application (found by Bayer and Mongardi) to the (almost complete) determination of the image of the period map for cubic fourfolds, hereby partially recovering a result of Laza.[-]