Indexed metadata

Nakano-positive determinants outside the Hodge-Riemann cone

Zhangchi Chen

Source record

Source: arXiv

Published: Sep 7, 2026

arXiv: 2609.07964

Open original source ↗

Source abstract

Dinh and Nguyên asked whether the determinant of a Griffiths positive matrix with (1,1)(1,1)-form entries belongs to the Hodge--Riemann cone. For every n4n\geqslant4 and 2kn22\leqslant k\leqslant n-2, we present an explicit Nakano positive k×kk\times k matrix of constant (1,1)(1,1)-forms on $\C^n$ whose determinant has a singular Lefschetz map in bidegree (1,nk1)(1,n-k-1). This gives a negative answer throughout this range. The boundary cases n=k4n=k\geqslant4 and n=k+15n=k+1\geqslant5 remain open. We next study the question under the simultaneous diagonalizability (SD) condition. Under this condition, we prove the Hodge--Riemann property in every bidegree (p,q)(p,q) with p+q=nkp+q=n-k and min(p,q)1\min(p,q)\leqslant1. Consequently, SD gives an affirmative answer when 1k=n21\leqslant k=n-2 or n3n-3. For every n6n\geqslant6 and 2kn42\leqslant k\leqslant n-4, however, we present SD examples whose Lefschetz map in bidegree (2,nk2)(2,n-k-2) is singular, showing that SD alone does not imply the Hodge--Riemann property in all bidegrees. The positive result uses the theory of dually Lorentzian polynomials developed by Ross, Süß, and Wannerer, in particular their generalized Alexandrov--Fenchel inequality and its equality characterization. Exact Python verification programs accompany the constructions.

Evidence graph

No public relationships recorded yet.

Integrity note: This page is a factual metadata record created by deterministic ingestion. It is not a claim that the work moves a mathematical frontier or has been independently verified.

Nakano-positive determinants outside the Hodge-Riemann cone — Mathematical Frontier Network