Higher Abel-Jacobi maps (Q1126727)

For the structure of the Chow group \(\text{CH}^p(X)\) of a smooth projective variety \(X\), there is a conjecture that there ought to be a filtration \(\text{CH}^p(X)\otimes \mathbb{Q}= F^0\text{CH}^p(X)\otimes \mathbb{Q}\supseteq F^1\text{CH}^p(X)\otimes \mathbb{Q}\supseteq\dots \supseteq F^{p+1}\text{CH}^p(X)\otimes \mathbb{Q}=0\) [see \textit{U. Jannsen}, in: Motives, Proc. Summer Res. Conf. Motives, Seattle 1991, Proc. Symp. Pure Math. 55, Part 1, 245-302 (1994; Zbl 0811.14004) for a discussion]. Concerning the first two terms, one has the cycle class map \(\psi_0: \text{CH}^p(X)\otimes \mathbb{Q}\to H^{2p}(X,\mathbb{Q})\cap H^{p,p}(X)\), conjecturally surjective by the Hodge conjecture, which should give rise to \(\text{ker} (\psi_0)= F^1\text{CH}^p(X)\otimes \mathbb{Q}\), and, by the work of Griffiths, one has the Abel-Jacobi map \(\psi_1: \text{ker}(\psi_0)\to J^p(X)\otimes \mathbb{Q}\) with \(\text{ker}(\psi_1)=F^2\text{CH}^p(X)\otimes \mathbb{Q}\). In this paper the author speculates about what comes after the Abel-Jacobi map, and constructs part of the missing map \(\psi_2\) in the case of 0-cycles on a surface, ``using a construction that has the potential to work more generally''.