The paper is about using tools from algebraic topology and algebraic geometry to study whole number and rational number solutions to polynomial equations. A central idea is that a geometric object called a variety (the solution set of polynomial equations) can be encoded in a rich algebraic structure. Points on the variety, meaning actual solutions, give rise to special maps called augmentations into this algebraic structure. The paper studies the collection of all such augmentations and asks whether this collection can itself be given the structure of a well-behaved geometric object, specifically a finite-type variety over the rational numbers.
The motivation comes from a program called Chabauty-Kim theory, which is a strategy for proving that certain equations have only finitely many whole number or rational number solutions by using the unipotent fundamental group, a sophisticated invariant that generalizes classical ideas from topology. The hope is to run this program in a more abstract, motivic setting without needing to resolve longstanding open problems in the foundations of the field. A parallel goal is to extract arithmetic information from higher-dimensional topological invariants of a space, going beyond just the fundamental group, which could give stronger tools for bounding the number of solutions.
The paper establishes two general criteria guaranteeing that the space of augmentations is representable as a geometric object of the desired kind. The authors compute this space explicitly in several examples, connect it to previously studied objects called Selmer varieties, and show how the construction leads to finiteness results phrased in terms of algebraic K-theory. The overall contribution is to put the collection of arithmetic points into a geometric framework that should make it more tractable to study using the tools of modern algebraic geometry.