Determine the generic automorphism group in the remaining signature cases

Determine the generic automorphism group of the universal abelian variety over the basic locus of a split unitary Shimura variety when the signature $(m_0,n_0-m_0)$ does not satisfy the hypotheses of Theorem 1.1, and establish whether it is larger than $\{\pm 1\}$.

Background

The main theorem proves that, under specified restrictions on the prime pp and the signature (m0,n0m0)(m_0,n_0-m_0), the generic automorphism group of the universal abelian variety on the basic stratum is exactly {±1}\{\pm1\}. The paper does not resolve the complementary signatures excluded from that theorem.

The authors explicitly expect the generic automorphism group in those remaining cases to be larger than {±1}\{\pm1\}. The remark also identifies a potentially relevant exceptional situation: when (g,p)=(2,2)(g,p)=(2,2) and m{1,n1}m\in\{1,n-1\}, the congruence subgroup VNV_N may contain nontrivial $2$-torsion, whose finite subgroups are expected to contribute groups of the form (Z/2Z)k(\mathbb Z/2\mathbb Z)^k.

References

If the signature $(m_0,n_0-m_0)$ is not as in the statement of Theorem \ref{thmmain1}, we expect the generic automorphism group of the universal abelian variety over the basic locus to be larger than ${\pm 1}$.

Oort's conjecture for split unitary Shimura varieties  (2609.04072 - Philippe et al., 3 Sep 2026) in Remark 6.1, “remaining cases”