We reference [EGNO] for the concept of a braided fusion category. Following the conventions in [JFR], let $\mathcal{C}$ denote a braided fusion category equipped with a braiding $\beta$, and let $\mathcal{B} \subset \mathcal{C}$ represent a fully faithful inclusion of braided fusion categories. The centralizer of this inclusion is defined as:

$$\mathcal{Z}_2 (\mathcal{B} \subset \mathcal{C}) := \{ y \in \mathcal{C} \ | \ \beta_{y,x} \circ \beta_{x,y} = {\rm id}_{x \otimes y} \ \forall x \in \mathcal{B} \}.$$

The Müger center $\mathcal{Z}_2 (\mathcal{B})$ of $\mathcal{B}$ is the centralizer of the identity inclusion. This center forms a symmetric fusion category. A braided fusion category $\mathcal{B}$ is termed nondegenerate if $\mathcal{Z}_2 (\mathcal{B})$ is equivalent to the trivial fusion category ${\rm Vec}$. A braided fusion category $\mathcal{B}$ is defined as slightly-degenerate if $\mathcal{Z}_2 (\mathcal{B})$ is equivalent to the fusion category ${\rm sVec}$ of $C_2$-graded finite-dimensional vector spaces. A slightly-degenerate braided fusion category $\mathcal{C}$ is referred to as split if it is equivalent to $\mathcal{D} \boxtimes {\rm sVec}$, where $\mathcal{D}$ is nondegenerate.

A fusion category $\mathcal{C}$ is called integral if the Frobenius-Perron dimension (${\rm FPdim}$) of each of its simple objects is an integer. An integral fusion category is inherently pseudo-unitary (i.e. ${\rm FPdim}(\mathcal{C}) = {\rm dim}(\mathcal{C})$), and consequently, spherical, as discussed in [EGNO]. A braided fusion category that possesses a spherical structure is (termed) premodular. If such a category is either nondegenerate or slightly-degenerate, it is denoted as modular or super-modular, respectively.

The classification of super-modular fusion categories began with the works cited in [BGHNPRW17, BGNPRW20, BPRZ21]. To date, very few non-split examples have been identified. The only known integral examples split as $\mathcal{D} \boxtimes \mathrm{sVec}^-$, which are non-positive (here positive means that ${\rm dim}(X) > 0$ for every simple object $X$). Perhaps the assumptions of integrality and positivity in a super-modular fusion category imply that it must necessarily split.

Question: Is there a non-split super-modular positive integral fusion category?


[BGHNPRW17] Bruillard, Paul; Galindo, César; Hagge, Tobias; Ng, Siu-Hung; Plavnik, Julia Yael; Rowell, Eric C.; Wang, Zhenghan. Fermionic modular categories and the 16-fold way. J. Math. Phys. 58 (2017), no. 4, 041704, 31 pp.
[BGNPRW20] Bruillard, Paul; Galindo, César; Ng, Siu-Hung; Plavnik, Julia Y.; Rowell, Eric C.; Wang, Zhenghan. Classification of super-modular categories by rank. Algebr. Represent. Theory 23 (2020), no. 3, 795--809.
[BPRZ21] Bruillard, Paul; Plavnik, Julia; Rowell, Eric C.; Zhang, Qing. On classification of super-modular categories of rank 8. J. Algebra Appl. 20 (2021), no. 1, Paper No. 2140017, 36 pp.
[EGNO] Etingof, Pavel; Gelaki, Shlomo; Nikshych, Dmitri; Ostrik, Victor. Tensor categories. Mathematical Surveys and Monographs, 205. American Mathematical Society, Providence, RI, 2015. xvi+343 pp.
[JFR] Johnson-Freyd, Theo; Reutter, David. Minimal nondegenerate extensions. J. Amer. Math. Soc. 37 (2024), no. 1, 81--150.
[Mu] Müger, Michael. On the structure of modular categories. Proc. London Math. Soc. (3) 87 (2003), no. 2, 291--308.

  • $\begingroup$ To make the question entirely self-contained (so rare and so nice!), could you please also explain what does "integral" mean? $\endgroup$ Apr 24 at 5:27
  • 1
    $\begingroup$ @მამუკაჯიბლაძე: Sure! Integral means that the Frobenius-Perron dimension of every simple object is an integer. I will add the definition. $\endgroup$ Apr 24 at 5:38

1 Answer 1


Example 2.1 in [BGHNPRW17] that corresponds to the centralizer of a fermion in the twisted Drinfeld double of $SL(2,5)$is an integral non-split super-modular category. In fact, if we use Proposition 3.2, then we can easily see (since $\mathbb A_5$ is simple) that the example does not contain modular categories of dimension 4; hence, by Theorem 3.1, it is non-split.

  • $\begingroup$ Since Z(Rep(A5)) is perfect, meaning it contains no non-trivial invertible objects, how can it possibly contains a fermion? $\endgroup$ Apr 24 at 13:13
  • $\begingroup$ Or maybe, you wanted to write "the Drinfeld double of SL(2,5)". $\endgroup$ Apr 24 at 16:20
  • $\begingroup$ In fact, a twisted Drinfeld double $D^{\omega}{\rm SL}(2,5)$ with $ord(\omega)=4$. $\endgroup$ Apr 24 at 16:55
  • $\begingroup$ Yes, I fixed the typo. $\endgroup$ Apr 24 at 17:21
  • $\begingroup$ To avoid any confusion, the example is $\mathcal{Z}_2 ({\rm sVec} \subset {\rm Rep}(D^{\omega}G))$, with $G={\rm SL}(2,5)$ and $ord(\omega)=4$. It is of course not the symmetric category ${\rm Rep}(G,z)$ first mentioned in your Example 2.1. $\endgroup$ Apr 24 at 17:43

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.