In this article, motivated by a problem asked by Allison and Panagiotopoulos, we study a problem concerning the complexity of group extensions within a hierarchy (denoted by -CLI and L--CLI) on the class of non-archimedean CLI Polish groups: Given a non-archimedean Polish group and one of its closed normal subgroup , suppose and are -CLI and -CLI, respectively. Is always -CLI? We provide a positive answer under a certain additional assumption. We then construct two examples yielding negative answers: for each countably infinite ordinal , there exists a group that is not -CLI, but has a -CLI normal subgroup such that is proper -CLI; there exists a proper -CLI group that has an abelian normal subgroup such that is also abelian. These examples also provide negative answers to the original problem raised by Allison and Panagiotopoulos. Finally, we show that if and are -CLI and -CLI with , respectively, then is -CLI, which gives an upper bound on the complexity of the extended group.

生成讲解博客的中文版本

生成AI智能校验的中文版本