Logarithmic Hesse's problem (Q420552): Difference between revisions

Take \(f\in \mathbb {C}[x_1,\dots ,x_n]\) and let \(H(f)\) be the Hessian of \(f\), i.e. the determinant of the Hesse matrix of \(f\). Contrary to a very old claim made by Hesse if \(n\geq 5\) there are homogeneous polynomials with \(H(f) \equiv 0\), but not depending from \(n-1\) variables, up to a change of coordinates [\textit{C. Lossen}, Bull. Braz. Math. Soc. 35, No. 1, 71--82 (2004; Zbl 1061.12002); \textit{C. Ciliberto}, \textit{F. Russo} and \textit{A. Simis}, Adv. Math. 218, No. 6, 1759--1805 (2008; Zbl 1144.14009)]. Here the author proves the surprising result that Hesse's claim is true for Laurent polynomials, i.e. for the coordinate ring of the \(n\)-torus \(\mathbb {C}^n\), i.e. if a Laurent polynomial \(f\) has logarithmic Hessian \(H(f)\equiv 0\), then \(f\) depends on at most \(n-1\) variables, up to an automorphism of \(\mathbb {C}^n\). He also defines and study the logarithmic polar map \(L_f\) of a Laurent polynomial map (\(L_f\) is not dominant if and only if \(H(f)\equiv 0\)).