The paper studies geometric invariants of certain spaces built from a base geometric object X. Specifically, it focuses on Chern-Schwartz-MacPherson (CSM) classes, which are a way of measuring the shape and topology of spaces that may have complicated singularities. For smooth spaces, these classes reduce to classical invariants coming from the tangent bundle, but CSM classes are powerful because they work even when the space is badly singular. The main computational achievement is working out these classes for "symmetric powers" of X, which are spaces formed by taking n copies of X and treating collections of n points as unordered. The results hold up to a specific type of ambiguity involving denominators related to the number n factorial.
Building on that computation, the authors calculate CSM classes for two important families of spaces sitting over symmetric powers: Quot schemes and Hilbert schemes of points. Hilbert schemes of points on X are spaces that parametrize finite collections of n points on X, but in a refined way that keeps track of algebraic data even when points collide. These spaces appear throughout modern algebraic geometry and mathematical physics. The key technique is pushing forward, meaning taking geometric information defined on a complicated space and systematically transferring it to a simpler one, in this case from the Hilbert scheme down to the symmetric power.
The final part of the paper connects to a famous open problem called the MNOP conjecture, which relates counts of curves in a three-dimensional geometric space to invariants coming from Hilbert schemes. The authors propose a sharper, more refined version of this conjecture that tracks where the fundamental class of the Hilbert scheme lands inside the symmetric power. They then prove this refined version, again up to the factorial ambiguity mentioned earlier, when X satisfies a special condition called being Calabi-Yau, a condition that arises naturally in string theory and mirror symmetry.