The paper deals with a concept called the "Bogomolov property," which is a way of saying that the smallest nontrivial algebraic numbers in a given collection of number fields cannot get arbitrarily close to 1 in a specific geometric sense. More precisely, for any field with this property, there is a positive lower bound on the "height" of algebraic numbers in that field, where height is a measure of arithmetic complexity. The central question is whether certain infinite collections of number fields, built by repeatedly extracting roots of numbers, inherit this property.
The specific fields under study are called radical extensions. You start with a base number field and a set of numbers inside it, then build a tower of new fields by adjoining all p-th roots of those numbers, for a chosen prime p. A conjecture by Rémond predicts that these large fields should have the Bogomolov property, except for a "trivial" region consisting of roots of the original numbers. In other words, numbers with very small height in these radical extensions should already come from the natural algebraic roots of the starting group, and nothing genuinely surprising should appear. The paper gives a new general criterion that guarantees this predicted behavior holds.
The authors then use this criterion to prove new cases of Rémond's conjecture, specifically for extensions built using division points relative to a finite set of primes. Their approach draws on a recent result about small points in so-called p-adic Lie extensions, which are fields with a rich symmetry structure described by p-adic analysis. By connecting the radical extension setting to this p-adic framework, they recover results previously obtained by other researchers and push them further, broadening the class of cases where Rémond's conjecture is now confirmed.