Powers of abelian fourfolds of Mumford type: a K3 criterion for the Hodge conjecture and explicit K3 partners

Abelian fourfolds A of Mumford type have Hodge rings generated by divisor classes, but A × A carries two exceptional Hodge classes of unknown algebraicity. First, the proof of Theorem 3.2 of Arapura's "Hodge cycles and the Leray filtration" (Pacific J. Math. 2022), the Hodge conjecture for fibre powers of Viehweg–Zuo families, has a gap for totally real fields of degree d ≥ 2, except in Mumford's case with one factor (noted earlier for two factors in an online thread). The theorem is then equivalent to the Hodge conjecture for a power of the very general fibre, known for d = 2 by an unrefereed preprint of Markman and a theorem of Abdulali and open for d ≥ 3 as far as we know. Second, if Galluzzi's K3-type sub-Hodge structure W(A) of H²(A × A) is realized by the transcendental lattice of a K3 surface S through an algebraic cycle, and the real multiplication of S is algebraic, then the Hodge conjecture holds for all powers of A and all products S^m × A^n. The proof, by twisted Casimir operators, extends to abelian varieties of Viehweg–Zuo type, where the Hodge conjecture for all powers reduces to the square. Third, for the Shimura families of genus-four curves of Bouchet, Hanselman, Pieper and Schiavone over Δ(2,3,7)\ℍ and Δ(2,3,9)\ℍ, explicit elliptic K3 surfaces S from families of van Geemen and Schütt satisfy T(S) ≅ W(A) for the Jacobian A at the same very general parameter, using results of that preprint whose proofs we have not checked. For them the Hodge conjecture for S × A² implies it for all S^m × A^k. An online thread had proposed a conditional link of this kind for the first family. For it we formulate the missing cycle as a half twist, a variant of a conjecture of Bouchet et al., with Frobenius-trace evidence. Closely related results for Mumford's fourfolds, including the reduction to the square, are in an unrefereed preprint of Bhattacharjee, Mandal and Bhattacharya; its proof for all powers of A assumes a withdrawn claim. Code and data are at github.com/sallerk/HC-mumford-squares.

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-14
DOI
https://doi.org/10.5281/zenodo.24214892
Primary Topic
Algebraic Geometry and Number Theory
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Powers of abelian fourfolds of Mumford type: a K3 criterion for the Hodge conjecture and explicit K3 partners

Claude Anthropic, K. Saller
Zenodo (CERN European Organization for Nuclear Research)
Algebraic Geometry and Number Theory
preprint

Powers of abelian fourfolds of Mumford type: a K3 criterion for the Hodge conjecture and explicit K3 partners

Claude Anthropic, K. Saller
preprint en

Abstract

Abelian fourfolds A of Mumford type have Hodge rings generated by divisor classes, but A × A carries two exceptional Hodge classes of unknown algebraicity. First, the proof of Theorem 3.2 of Arapura's "Hodge cycles and the Leray filtration" (Pacific J. Math. 2022), the Hodge conjecture for fibre powers of Viehweg–Zuo families, has a gap for totally real fields of degree d ≥ 2, except in Mumford's case with one factor (noted earlier for two factors in an online thread). The theorem is then equivalent to the Hodge conjecture for a power of the very general fibre, known for d = 2 by an unrefereed preprint of Markman and a theorem of Abdulali and open for d ≥ 3 as far as we know. Second, if Galluzzi's K3-type sub-Hodge structure W(A) of H²(A × A) is realized by the transcendental lattice of a K3 surface S through an algebraic cycle, and the real multiplication of S is algebraic, then the Hodge conjecture holds for all powers of A and all products S^m × A^n. The proof, by twisted Casimir operators, extends to abelian varieties of Viehweg–Zuo type, where the Hodge conjecture for all powers reduces to the square. Third, for the Shimura families of genus-four curves of Bouchet, Hanselman, Pieper and Schiavone over Δ(2,3,7)\ℍ and Δ(2,3,9)\ℍ, explicit elliptic K3 surfaces S from families of van Geemen and Schütt satisfy T(S) ≅ W(A) for the Jacobian A at the same very general parameter, using results of that preprint whose proofs we have not checked. For them the Hodge conjecture for S × A² implies it for all S^m × A^k. An online thread had proposed a conditional link of this kind for the first family. For it we formulate the missing cycle as a half twist, a variant of a conjecture of Bouchet et al., with Frobenius-trace evidence. Closely related results for Mumford's fourfolds, including the reduction to the square, are in an unrefereed preprint of Bhattacharjee, Mandal and Bhattacharya; its proof for all powers of A assumes a withdrawn claim. Code and data are at github.com/sallerk/HC-mumford-squares.

Zenodo (CERN European Organization for Nuclear Research)
Algebraic Geometry and Number Theory
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon
Powers of abelian fourfolds of Mumford type: a K3 criterion for the Hodge conjecture and explicit K3 partners — Claude Anthropic, K. Saller · Zenodo (CERN European Organization for Nuclear Research) (2026) | TGRS Research Map | TGRS