Balanced tensor categories of representations of fixed-points conformal nets

This paper establishes an equivalence of balanced $\mathrm{W}^*$-tensor categories between the $G$-equivariantization of the category of $G$-twisted representations of a conformal net $\mathcal{A}$ and the category of representations of its fixed-point net $\mathcal{A}^G$, thereby extending a known rational result to the non-rational case while preserving the balanced structure.

The Gist

Imagine the universe of physics as a giant, intricate tapestry woven from invisible threads of energy and symmetry. In the world of Conformal Field Theory (CFT), mathematicians use a tool called a Conformal Net to map out these threads. Think of a Conformal Net as a sophisticated instruction manual that tells you how to build and manipulate these energy threads on a circle (representing a slice of time and space).

This paper, written by Adrià Marín-Salvador, tackles a specific puzzle in this mathematical universe: What happens when you take a complex system and force it to obey a strict set of rules (symmetries)?

Here is the story of the paper, broken down into simple concepts and analogies.

1. The Setup: The Original System and the "Orbifold"

Imagine you have a massive, chaotic dance floor (the Conformal Net, let's call it A). Dancers (representations) move around, following complex rules.

Now, imagine a group of strict choreographers (a Finite Group G) arrives. They demand that the dance floor must look the same no matter how they spin or flip the room. They enforce a rule: "If you rotate the room, the dance must look identical."

When you apply these rules, you don't just get a smaller dance floor; you get a Fixed-Points Net (A_G). This is the new, simplified version of the system where only the moves that survive the choreographers' scrutiny remain.

The Big Question: If we know all the possible dances on the original floor (A), can we predict all the possible dances on the new, restricted floor (A_G)?

2. The Problem: Missing Pieces

In the past, mathematicians knew the answer for "simple" dance floors (called Rational systems). They found a perfect dictionary to translate dances from the old floor to the new one.

However, most real-world systems are not simple. They are messy, with infinite variations and continuous flows of energy. The old dictionary broke down for these complex systems. The paper asks: Can we build a new dictionary that works for the messy, complex systems too?

3. The Solution: Twisted Representations and "Equivariantization"

To solve this, the author introduces two clever concepts:

• Twisted Representations (The "Disguised" Dancers):
  In the original system, some dancers don't just follow the rules; they follow the rules with a twist. Imagine a dancer who, every time they pass a specific point on the circle, secretly swaps their costume according to the choreographer's instructions. These are Twisted Representations.
  The paper shows that to understand the new, restricted floor (A_G), you can't just look at the normal dancers. You must collect all the normal dancers and all the twisted dancers together.

• Equivariantization (The "Team-Up" Process):
  Once you have gathered all the normal and twisted dancers, you have a huge, chaotic pile. The paper introduces a process called Equivariantization. Think of this as a "team-building exercise."
  You take this pile of dancers and force them to form teams where every member agrees on the choreographer's rules. You filter out the chaos and organize the twisted dancers into a structured group that respects the symmetry.

4. The Main Discovery: The Perfect Match

The paper's main result is a mathematical "Aha!" moment. It proves that:

The collection of all dances on the new, restricted floor (A_G) is exactly the same as the organized team of normal and twisted dancers from the old floor.

In mathematical terms, the category of representations of the fixed-point net is equivalent to the equivariantization of the category of twisted representations.

The Analogy:
Imagine you have a giant library of books (the original system). Some books are standard, and some are "twisted" (written in a code that changes based on the reader).

• The Old Way: You tried to find the "Fixed-Point Library" (the books that make sense under strict rules) by only looking at the standard books. It didn't work.
• The New Way: The author says, "Gather all the books, including the coded ones. Then, organize them into a 'Symmetry Club' where every book agrees on the rules."
• The Result: The "Symmetry Club" you created is identical to the "Fixed-Point Library." You didn't lose anything, and you didn't gain anything extra; you just found the right way to organize the pieces.

5. Why This Matters (In the Paper's Context)

The paper doesn't just say "they are the same." It proves they are the same in a very specific, high-level way:

• Balanced: The paper ensures that the "twist" or "balance" (a mathematical property related to how things rotate and braid) is preserved perfectly during the translation.
• General: It works even when the system is messy and infinite (non-rational), not just when it's simple and finite.

Summary

This paper is like finding a universal translator for a complex language. It proves that if you want to understand a system that has been stripped down by symmetry rules, you don't need to start from scratch. Instead, you can take the original system, add in the "twisted" versions of its parts, organize them into a coherent group, and you will get a perfect, one-to-one match with the simplified system.

The author achieves this by building a bridge using Connes fusion (a way of gluing mathematical objects together) and proving that this bridge holds up even for the most complex, non-rational systems. It generalizes a known result from simple systems to the messy, real-world-like systems, ensuring the mathematical "balance" remains intact throughout the process.

Technical Summary

Technical Summary: Balanced Tensor Categories of Representations of Fixed-Points Conformal Nets

Problem Statement
Conformal nets provide a rigorous mathematical framework for two-dimensional chiral conformal field theory (CFT). A central problem in the theory is understanding the representation theory of a "fixed-points" conformal net $A^G$, obtained by modding out a conformal net $A$ by a faithful action of a finite group $G$ (orbifolding). While the relationship between the representations of $A$ and $A^G$ is well-understood in the rational case (where the category of representations is a modular tensor category), the non-rational case presents significant challenges. In the non-rational setting, the category of representations $\text{Rep}(A)$ is generally not finite, semisimple, or rigid; it involves direct integrals of irreducible representations rather than direct sums.

Previous work by Müger established an equivalence of braided tensor categories $(G\text{-Loc}(A))^G \cong \text{DHR}(A^G)$ for rational nets, where $G\text{-Loc}(A)$ is the category of $G$-localized endomorphisms. More recently, the author established that the category of $G$-twisted representations, $\text{Rep}_G(A)$, forms a $G$-crossed balanced $W^*$-tensor category. However, a comprehensive equivalence for the non-rational case that preserves the "balance" (categorical twist) structure was lacking. This paper addresses the gap by generalizing the Müger equivalence to non-rational conformal nets and upgrading the result to an equivalence of balanced $W^*$-tensor categories.

Methodology
The paper employs the language of $W^*$-tensor categories and Connes fusion of representations, avoiding the localized endomorphism language used in earlier rational results. The methodology proceeds through the following steps:

1. Framework of $G$-Crossed Balanced Categories: The author utilizes the previously established structure of $\text{Rep}_G(A)$ (the direct sum of categories of $g$-twisted representations for $g \in G$) as a $G$-crossed balanced $W^*$-tensor category. This includes a $G$-grading, a $G$-action by tensor automorphisms, a $G$-crossed braiding, and a $G$-crossed balance.
2. G-Crossed Categorical Extensions: To bridge the gap between the twisted representations and the fixed-point net, the paper introduces the notion of a "$G$-crossed categorical extension" of a conformal net. This concept generalizes the categorical extensions defined by Gui. The author proves a uniqueness theorem (Theorem 3.5) stating that any two such extensions of a conformal net $A$ are equivalent.
3. Construction of the Module Category: The vacuum representation $H_0$ of $A$, when restricted to the fixed-point net $A^G$, is shown to carry the structure of a commutative separable $C^*$-Frobenius algebra object, denoted $\Gamma$, in $\text{Rep}(A^G)$. The author considers the category $\text{Mod}(\Gamma)$ of unitary modules over $\Gamma$.
4. Equivalence via De-equivariantization: The core strategy involves showing that $\text{Mod}(\Gamma)$ is equivalent to $\text{Rep}_G(A)$ as a $G$-crossed braided $W^*$-tensor category. This is achieved by constructing a functor $M: \text{Mod}(\Gamma) \to \text{Rep}_G(A)$ and utilizing the uniqueness of $G$-crossed categorical extensions to upgrade this to an equivalence.
5. Equivariantization: The paper then considers the $G$-equivariantization of $\text{Mod}(\Gamma)$, denoted $\text{Mod}(\Gamma)^G$. A functor $D: \text{Rep}(A^G) \to \text{Mod}(\Gamma)^G$ is constructed, which is shown to be an equivalence of braided $W^*$-tensor categories.
6. Compatibility with Balance: Finally, the author verifies that the equivalence respects the balance (categorical twist) structures on both sides, ensuring the result holds for balanced tensor categories.

Key Contributions and Results
The primary result of the paper is Theorem 4.19, which establishes the following equivalence:
$$\text{Rep}(A^G) \cong (\text{Rep}_G(A))^G$$
where:

• $A$ is a conformal net (not necessarily rational) acted on faithfully by a finite group $G$.
• $\text{Rep}(A^G)$ is the category of representations of the fixed-point conformal net.
• $\text{Rep}_G(A)$ is the $G$-crossed balanced $W^*$-tensor category of $G$-twisted representations of $A$.
• $(\text{Rep}_G(A))^G$ is the $G$-equivariantization of $\text{Rep}_G(A)$.

This equivalence is an equivalence of balanced $W^*$-tensor categories. The underlying functor $R: (\text{Rep}_G(A))^G \to \text{Rep}(A^G)$ restricts a $G$-equivariant twisted representation on a Hilbert space $H$ to the honest representation of $A^G$ on the subspace of $G$-invariant vectors.

Specific technical contributions include:

• Generalization to Non-Rational Nets: The result removes the rationality hypothesis required in Müger's original theorem, accommodating categories with direct integrals and continuously many simple objects.
• Preservation of Balance: The paper explicitly constructs and verifies the compatibility of the equivalence with the balance (categorical twist), a structure often omitted in non-rational contexts.
• Uniqueness of Extensions: Theorem 3.5 provides a uniqueness result for $G$-crossed categorical extensions, which is instrumental in proving the equivalence between the module category $\text{Mod}(\Gamma)$ and the twisted representation category $\text{Rep}_G(A)$.
• Involutive Structures: The paper notes that the equivalence can be upgraded to be compatible with involutive structures (Appendix A), as constructed in a companion work.

Significance
The paper claims that these results provide the first explicit computation of a category of representations for a conformal net where irreducible representations tensor to a direct integral of irreducible representations, rather than a direct sum. By generalizing the Müger equivalence to the non-rational setting and including the balance, the work extends the structural understanding of orbifolding in CFT beyond the realm of modular tensor categories. The methodology relies on the interplay between Connes fusion, categorical extensions, and the theory of Frobenius algebras in $W^*$-tensor categories, offering a robust framework for analyzing fixed-point nets without assuming semisimplicity or finiteness.