On the Shafarevich and Tate conjectures for hyperkähler varieties (Q1923227): Difference between revisions

In this article, the author considers the following 3 important problems: (1) (Shafarevich problem). Are there only finitely many polarized \(K3\) surfaces of fixed degree \(d\) over a number field \(K\), with good reduction outside a fixed finite set of primes? (2) Describe the motive of a \(K3\) surface, and compute the motivic Galois group. (3) (Tate problem) Let \(Y\) and \(Y'\) be \(K3\) surfaces defined over a number field \(K\). Is any isomorphism of \(\operatorname {Gal}(\overline{K}/K)\)-modules \(H_{\text{et}}^*(Y_{\overline K}, {\mathbb{Q}}_{\ell}) \rightarrow H_{et}^*(Y'_{\overline K}, {\mathbb{Q}}_{\ell})\) induced by a \({\mathbb{Q}}_{\ell}\)-linear combination of algebraic cycles on \(Y \times Y'\)? Is the image of \(\operatorname {Gal}{\overline K}/K)\) in \(GL H_{\text{et}}^*(Y_{\overline K}, {\mathbb{Q}}_{\ell})\) as big as possible, i.e., an open Lie subgroup of the motivic Galois group over \({\mathbb{Q}}_{\ell}\)? The author gives a positive answer to problem (1) and corollary (just name one here): For any positive integer \(m\), there are only finitely many orbits for \(PGL(6, {\mathbb{Z}}[1/3m])\) among all smooth cubics in the \(5\)-dimensional projective space over \({\mathbb{Z}}[1/3m]\). Also the problem (2) is answered and a partial answer to problem (3) is given in the paper, both for not just \(K3\) surfaces but also for polarized varieties satisfying axioms \(A_k\), \(B_k\) or \(B_k^+\); one notes that polarized abelian surfaces, surfaces of general type with \(p_g = 1\) and \(K^2 = 1\), \(K3\) surfaces and hyperkahler varieties with \(b_2 > 3\) satisfy \(A_1, B_1^+\), while cubibic 4-folds satisfy \(A_2, B_2^+\).