Fusion Higher-Categories Overview

• Fusion higher-categories are higher-dimensional generalizations of fusion 1-categories, characterized by semisimplicity, rigidity, and fully dualizable objects.
• They integrate concrete constructions like 2Vect and bicommutant categories to connect quantum algebra with operator-algebraic frameworks.
• Their classification via Morita theory and higher Tannakian data offers deep insights for topological field theory and quantum invariants.

Fusion higher-categories generalize the concept of fusion 1-categories—semisimple rigid tensor categories with finitely many simple objects—to higher categorical settings, particularly 2-categories and beyond. This framework synthesizes ideas from quantum algebra, topological field theory, von Neumann algebras, and higher representation theory. The subject includes concrete constructions, classification theorems, and categorical analogues of operator-algebraic phenomena, especially through the theory of bicommutant categories and their Morita theory.

1. Foundations and Basic Structures

A fusion 2-category is a $\Bbbk$-linear, finite semisimple, rigid monoidal 2-category with finitely many simple objects (up to equivalence), every object and 1-morphism is fully dualizable, and the unit object is simple. This generalizes the notion of fusion 1-categories, such as representation categories of finite groups or semisimple quantum groups at root of unity, to a 2-categorical context (Décoppet et al., 2023, Décoppet, 2022).

Key examples include:

• 2Vect: The 2-category of finite semisimple (perfect) $\Bbbk$-linear categories under the Deligne tensor product.
• 2Vect$_G$: $G$-graded 2-vector spaces, with a twist by a normalized 4-cocycle $\pi\in Z^4(G;\Bbbk^\times)$, yielding $2Vect_G^\pi$.
• 2Rep$(\mathcal{G})$: 2-representations of finite 2-groups.
• Module 2-categories $\mathbf{Mod}(\mathcal{B})$: Module 2-categories over a braided fusion 1-category.

Fusion $n$-categories (for general $n$) are expected to exhibit an inductive structure, with the "loop" category $\Bbbk$0 (endomorphisms of the unit) being a braided fusion $\Bbbk$1-category, and their centers classified via higher Tannakian theory (Décoppet et al., 2024).

2. Bicommutant Categories and Positivity in Higher Categories

Bicommutant categories are higher categorical analogues of von Neumann algebras. For a unitary fusion category $\Bbbk$2, any fully faithful unitary tensor functor $\Bbbk$3 (bimodules over a hyperfinite factor $\Bbbk$4) gives rise to its commutant $\Bbbk$5 and bicommutant $\Bbbk$6, categorifying the double-commutant theorem. The bicommutant category is defined as a positive bi-involutive category $\Bbbk$7 equipped with a suitable representation into bimodules so that the canonical inclusion $\Bbbk$8 is an equivalence (Henriques et al., 2020, Henriques et al., 2015).

A bi-involutive tensor category is a dagger tensor category endowed with conjugation (an antilinear anti-tensor functor), natural unitary isomorphisms $\Bbbk$9, $_G$0, and $_G$1 satisfying coherence. Positivity is encoded as a structural cone $_G$2 ("cp-maps") in the morphism spaces, with stability under composition, adjoint, and amplification. In the operator-algebraic case, these structures align with cp-maps and positive operators in von Neumann algebra theory.

In the fusion context, all unitary fusion categories are shown to admit faithful representations such that they are bicommutant categories. This result provides a higher-categorical bridge between fusion categories and von Neumann algebra theory, with direct implications for the study of local operator algebras in topological field theory (Henriques et al., 2015, Henriques et al., 2020).

3. Fusion 2-Categories: Internal Structure and Classification

A fusion 2-category is a rigid, semisimple monoidal 2-category with simple unit and finitely many simple objects. The theory of rigid and separable algebras within fusion 2-categories underpins their module and bimodule structures (Décoppet, 2022).

A rigid algebra $_G$3 in a fusion 2-category $_G$4 is specified by a multiplication $_G$5 admitting a right adjoint $_G$6 in the 2-category of $_G$7–$_G$8-bimodules, together with bimodule compatibility conditions on the adjunction data. $_G$9 is separable if the counit $G$0 splits as a bimodule morphism. A fundamental result is that a connected rigid algebra $G$1 is separable if and only if its categorical dimension in $G$2 is nonzero.

The module and bimodule 2-categories over separable algebras in a fusion 2-category inherit finite semisimplicity, echoing classical results in 1-category theory, and forming the backbone for the Morita theory of fusion 2-categories (Décoppet, 2022).

Classification of fusion 2-categories is achieved via a homotopy-coherent groupoid-theoretic framework (Décoppet et al., 2024). There is an equivalence between the 3-groupoid of multi-fusion 2-categories and a groupoid of commuting squares of $G$3-equivariant spaces, involving braided fusion categories, group inclusions, and cohomological data. In the bosonic case, the data comprise a pair of groups $G$4, a 4-cocycle $G$5, and a braided action; in the fermionic case, analogous super-cohomological data is used.

4. Morita Theory and Module 2-Categories

Morita theory in the 2-categorical setting identifies when two fusion 2-categories are "the same" from the perspective of module 2-categories or bimodule equivalence. The Morita 3-category of separable algebras in a fusion 2-category encodes these relationships, with 1-morphisms as bimodules, 2-morphisms as bimodule functors, and 3-morphisms as bimodule natural transformations (Décoppet, 2022).

Morita equivalence between fusion 2-categories is characterized by:

1. Equivalence of their 3-categories of separable module 2-categories.
2. Existence of a separable, faithful left module 2-category $G$6 such that $G$7.
3. Existence of a separable algebra $G$8 in $G$9 with $\pi\in Z^4(G;\Bbbk^\times)$0 as monoidal 2-categories.

Dual tensor 2-categories (endo-2-functors of a module 2-category) generalize the categorical centers and are shown to be multifusion 2-categories under mild separability conditions (Décoppet, 2022).

5. Group-Theoretical and Exotic Fusion 2-Categories

Group-theoretical fusion 2-categories arise as categorifications of group-theoretical fusion 1-categories. They are precisely those for which the category of endomorphisms of the unit object ($\pi\in Z^4(G;\Bbbk^\times)$1) is Tannakian, i.e., equivalent to $\pi\in Z^4(G;\Bbbk^\times)$2 for some finite group $\pi\in Z^4(G;\Bbbk^\times)$3 (Décoppet et al., 2023).

Explicitly, given finite groups $\pi\in Z^4(G;\Bbbk^\times)$4, $\pi\in Z^4(G;\Bbbk^\times)$5, a 4-cocycle $\pi\in Z^4(G;\Bbbk^\times)$6 and a 3-cochain $\pi\in Z^4(G;\Bbbk^\times)$7 with $\pi\in Z^4(G;\Bbbk^\times)$8, one defines the fusion 2-category $\pi\in Z^4(G;\Bbbk^\times)$9. Morita and monoidal equivalence of such categories are classified via group isomorphisms and twisting by 3-cochains in the expected cohomological manner.

Further, Tambara–Yamagami fusion 2-categories, classified in (Décoppet et al., 2023), are constructed via $2Vect_G^\pi$0-graded data, with the nontrivial component equivalent to $2Vect_G^\pi$1. Their equivalence classes are determined by pairs $2Vect_G^\pi$2, where $2Vect_G^\pi$3 is an abelian group and $2Vect_G^\pi$4 with specific constraints. These models support and categorize defects and duality phenomena observable in 2+1-dimensional topological phases.

6. Structural Theorems and Rigidity

The classification of fusion 2-categories demonstrates strong analogues to Ocneanu rigidity found in 1-categorical fusion theory. Namely, for any fixed invariants (rank, underlying fusion 1-category), there are only finitely many fusion 2-categories up to equivalence (Décoppet et al., 2024). Furthermore, fusion 2-categories admit no nontrivial deformations: the group of monoidal autoequivalences is finite and their infinitesimal deformation space vanishes, mirroring the rigorous constraints of Tannakian theory and categorical cohomology.

The fusion rules of group-theoretical fusion 2-categories admit explicit descriptions in terms of double cosets and module categories over intersections of subgroups twisted by explicit cocycles (Décoppet et al., 2023). This structure determines the "hypergroup" of components and supports the inductive program for higher $2Vect_G^\pi$5-categorical analogues (Décoppet et al., 2024).

7. Outlook: Higher Fusion Categories and Connections

The pattern established for fusion 2-categories suggests that fusion $2Vect_G^\pi$6-categories can be approached inductively: with "unit endomorphisms" providing a braided fusion $2Vect_G^\pi$7-category, higher Tannakian data yielding group or supergroup extensions, and higher cohomological invariants (such as $2Vect_G^\pi$8) encoding associativity and anomaly structures. This framework interfaces directly with the cobordism hypothesis, fully extended TQFT, and the study of defects and duality in quantum topological phases, providing a unified, group-theoretic, and cohomological landscape for fusion in higher categories (Décoppet et al., 2024).