← Back to arXiv
arXivAlgebraic GeometryarXiv:2607.18322

Picard groups and composition nilpotence for finite cellular isotropic spectra

The paper works in a branch of mathematics called motivic homotopy theory, which studies geometric objects (algebraic varieties) using tools borrowed from topology. A central player is the "stable motivic homotopy category" over a field, which packages geometric information in a way that allows algebraic and topological techniques to interact. The authors focus on a special slice of this world called "finite cellular isotropic spectra," built over flexible fields (a technical condition ensuring the field has lots of square roots available). The main computational tool is a cohomology theory denoted E, constructed by combining two natural objects in the theory.

The first major result classifies all "Picard group" elements in this setting. The Picard group collects the objects that behave like units under the tensor product operation, meaning objects you can multiply together and get back something equivalent to the baseline object. The authors prove this group is generated by just two independent integer parameters, corresponding to the two "directions" of suspension available in the motivic setting. In plain terms, every invertible finite cellular isotropic spectrum is simply a shifted version of one fundamental object, the isotropic sphere, with no exotic surprises. This is a clean and satisfying structural result, since Picard groups in motivic homotopy theory are often complicated and poorly understood.

The second major result concerns "composition nilpotence," a property saying that if you compose a map with itself enough times, you eventually get zero. The authors show that for any nonzero finite cellular object, the maps that become invisible to the cohomology theory E form a nilpotent ideal, and they give an explicit upper bound on how many times you need to compose before hitting zero. This bound depends on two numerical measurements of the object's complexity. A consequence is that E detects all the essential structural decompositions of these objects, providing a powerful computational handle on the category that was not available before.

Read original →