Incompressible tensor categories

Abstract

A symmetric tensor category D over an algebraically closed field k is incompressible if every tensor functor out of D is an embedding. E.g., the categories Vec and sVec of (super)vector spaces are incompressible. Moreover, by Deligne's theorem, if char(k)=0 then any tensor category of moderate growth uniquely fibres over sVec, so Vec and sVec are the only incompressible categories in this class. Similarly, in characteristic p>0, we have the incompressible Verlinde category Verp, and any Frobenius exact category of moderate growth uniquely fibres over Verp. More generally, the Verlinde categories Verpn, Verpn+ are incompressible, and a key conjecture is that every tensor category of moderate growth uniquely fibres over Verp∞. This would make the above the only incompressible categories in this class. We prove a part of this conjecture, showing that every tensor category of moderate growth fibres over an incompressible one. So it remains to understand incompressible categories. We say that D is subterminal if it every tensor category admits at most one fibre functor to it, and a Bezrukavnikov category if the class of tensor categories that fibre over D is closed under quotients. Clearly, a subterminal Bezrukavnikov category is incompressible, and we conjecture the converse. We prove that Verp is Bezrukavnikov, generalizing the result of Bezrukavnikov for Vec. We also find intrinsic sufficient conditions for incompressibility and subterminality. Namely, D is maximally nilpotent if the growth rates of symmetric powers are minimal. We show that a finite maximally nilpotent category is incompressible, and also subterminal if it satisfies an additional geometric reductivity condition. Then we verify these conditions for Ver2n.

0

Turn this paper into a lesson

ArcXiv compiles a structured reading guide from this paper's metadata: plain-English importance, contributions, prerequisite concepts, which sections to read first, flashcards, and a quiz. Grounded in the abstract, never invented.

Discussion (0)

Sign in to join the discussion.

Loading comments…