Algebraic geometers often want to classify geometric objects, like curves or surfaces, by constructing a "moduli space" whose points correspond to the objects being classified. Sometimes the most natural construction produces a more elaborate object called a stack, which keeps track of symmetries that a plain space would ignore. A classical result called the Keel-Mori theorem says that any well-behaved stack with finite symmetry groups (technically, finite inertia) can be compressed down into an ordinary space called a coarse moduli space, which captures the same classification data but forgets the symmetry bookkeeping.
Supergeometry is a mathematical framework that extends ordinary geometry by introducing extra "odd" or anticommuting coordinates, motivated originally by supersymmetry in physics. The objects of study are superspaces and superstacks, which mix ordinary geometric structure with these anticommuting directions. Extending classical theorems to this setting is nontrivial because the odd coordinates behave quite differently from ordinary ones, and standard algebraic tools need to be carefully adapted or replaced.
The paper proves that the Keel-Mori theorem holds in this super setting: every sufficiently nice superstack (specifically a Deligne-Mumford superstack with finite inertia, defined over a base where 2 is invertible) has a coarse moduli superspace. The requirement that 2 be invertible is a technical condition related to how the odd coordinates interact with characteristic 2 arithmetic. To get there, the authors develop new tools, including a notion of "integral morphisms" in supergeometry and a descent result that controls how geometric data can be glued together. They also show the result is sharp by exhibiting a superstack that falls just outside the assumed conditions and genuinely fails to have a coarse moduli superspace.