The Nash problem for a toric pair and the minimal log-discrepancy (Q710844): Difference between revisions

Given a variety \(X\), the Nash problem on arc spaces asks if the set of Nash components and the set of essential divisors over \(X\) are in bijection. It has a positive answer for toric varieties [\textit{S. Ishii} and \textit{J. Kollár}, Duke Math. J. 120, No. 3, 601--620 (2003; Zbl 1052.14011)]. The paper under review is considering the Nash problem for toric pairs \((X,Z)\), consisting of a toric variety and a \(\mathbb{T}\)-invariant closed subscheme [\textit{P. Petrov}, Math. Nachr. 282, No. 11, 1575--1583 (2009; Zbl 1182.14008)], as a particular case of more general Nash problem for pairs, satisfying two conditions. In particular, it should be possible to have some information about \(\text{sing}(X)\) from its Nash components in the space of arcs, and the main result of the paper is an example. It claims that for a toric pair \((X,Z)\) the problem has affirmative answer, and, if \(X\) is \(\mathbb{Q}\)-Gorenstein and \((X, Z)\) log canonical, any log Nash divisor computes its minimal log discrepancy, while if the pair is not log canonical, there is a log Nash divisor with negative log discrepancy (so that \(mld(X,Z)= -\infty\)).