The paper develops a mathematical framework for studying certain geometric objects called quasimaps, which are a generalization of maps from curves into algebraic varieties. Ordinary maps must send every point of the curve to a well-defined point of the target space, but quasimaps are allowed to be undefined at finitely many points. The authors build moduli spaces, which are geometric spaces that parametrize all quasimaps of a given type, much like how a moduli space of triangles would parametrize all triangles up to some equivalence. A key feature of their work is that they allow two separate parameters to vary simultaneously: how heavily the boundary of the target space is weighted, and how much the quasimap is penalized for being undefined at points. By varying these parameters, one can move between different regimes with different notions of stability.
As the parameters change, the moduli spaces themselves change, and the paper carefully tracks these changes through what are called wall crossings. At certain special parameter values, one regime transitions to another, and the authors construct explicit maps between the corresponding moduli spaces. A central technical piece is a new type of moduli space in the so-called Calabi-Yau regime, which sits between two other natural regimes called stable and log Fano. This intermediate space is subtle to construct because the stability condition there is numerically borderline, but the authors handle it by retaining extra data in the form of a polarization, essentially a choice of line bundle that helps distinguish objects that would otherwise look the same.
In the special case where the boundary is a single divisor of degree one, the abstract theory becomes very concrete: it recovers and connects classical objects like Hassett's spaces of weighted stable curves and geometric invariant theory quotients of weighted points on the projective line. These are well-studied compactifications of spaces of point configurations, and seeing them appear as special cases validates the general framework. A companion paper uses this machinery to connect two very different-looking compactifications of the space of rational elliptic surfaces, showing that what seemed like unrelated constructions are actually linked through the wall crossing theory developed here.