Beilinson-Hodge cycles on semiabelian varieties (Q730672)

Let \(U/{\mathbb C}\) be a smooth complex quasiprojective variety, and \(CH^i(U,j)\) the higher Chow group introduced by S. Bloch. Concerning the cycle class map \[ cl_{i,j} : CH^i(U,j)\otimes{\mathbb Q} \to \hom_{\text{MHS}}({\mathbb Q}(0),H^{2i-j}(U,{\mathbb Q}(i))), \] Beilinson conjectured that \(cl_{i,j}\) is surjective, where the case \(j=0\) amounts to the statement of the Hodge conjecture. It was Jannsen who first found a counterexample in the case \(i > j=1\), and since then other counterexamples where found by the reviewer and \textit{M. Kerr} [Invent. Math. 170, No. 2, 355--420 (2007; Zbl 1139.14010)]. Anecdotal evidence based on examples in the works of M. Asakura and S. Saito suggests that this conjecture holds in the case \(i=j\), which is now coined the Beilinson-Hodge conjecture. In the paper under review, the authors give an interesting proof of the Beilinson-Hodge conjecture for \(U\) being a product of curves or a semiabelian variety. Their method is based on a study of invariants under the Mumford-Tate group, reducing to the situation where \(i=j=1\), where the conjecture is easily known to hold.