🔢 mathematics

Lifts of symmetries to minimal nondegenerate extensions are unique up to fermion attachment

This paper provides a direct tensor-categorical proof that lifting symmetries from a slightly degenerate braided fusion category to a fixed minimal nondegenerate extension yields exactly two distinct locality-respecting classes related by fermion attachment, characterized by a surjective restriction homomorphism with a central $\mathbb{Z}/2$ kernel.

Original authors: Ran Luo, Jiahua Tian

Published 2026-10-14

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Ran Luo, Jiahua Tian

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

In the invisible world of quantum materials, there exists a strange kind of order that defies the usual rules of solid matter. Imagine a substance where the particles do not just sit still or flow like a liquid, but instead carry a hidden memory of how they have moved around one another. This is the realm of topological order, a state of matter found in certain exotic materials where the properties of the whole system depend on the global shape of the paths particles take, rather than just their local positions. Within this world, there are two distinct types of particles that physicists have learned to distinguish: ordinary particles that behave like standard matter, and a special kind of particle called a fermion, which acts as a local building block with a unique rule: if you swap two of them, the entire system flips its sign, a subtle but profound change in its quantum state.

When scientists study these materials, they often look for symmetries, which are ways the system can be transformed without changing its essential nature. However, the presence of these special fermions creates a puzzle. If you try to remove the fermions from the picture to see the underlying structure, or if you try to combine the system with others, the rules for how symmetries behave become ambiguous. Specifically, when researchers attempt to "gauge" the fermion parity—a process that effectively turns the local fermion into a new type of force carrier—they find that the symmetries of the original system can be lifted to the new, larger system in more than one way. The question that has lingered is whether this ambiguity is a small, manageable detail or a fundamental feature that allows for many different possibilities.

A team of researchers at Peking University and East China Normal University has now provided a definitive answer to this question. They have proven that for a wide class of these quantum systems, the ambiguity is exactly as large as it can possibly be, but no larger. When a symmetry of the original system is lifted to the new system created by gauging the fermion, there are precisely two distinct ways to do it. These two ways are not random variations; they are locked together by a single, specific action. One version of the symmetry acts normally, while the other version attaches a copy of the local fermion to every single "vortex" or defect in the new system. The researchers showed that these are the only two options available. There is no third way, and there is no way to twist the symmetry into something entirely different. The two possibilities are related by a simple, binary choice: either you leave the vortices alone, or you bind the fermion to them.

To reach this conclusion, the team used a rigorous mathematical framework that treats these quantum systems as networks of relationships rather than just collections of particles. They focused on the "vortex sector," the part of the new system where the fermions are no longer visible as simple particles but appear as twists in the fabric of the material. They demonstrated that any symmetry acting on this sector must either do nothing to the vortices or attach the fermion to all of them. They then proved that if a symmetry attaches the fermion, it cannot be confused with the symmetry that does nothing, even if they look identical on the surface. This is because the act of attaching the fermion introduces a specific sign change—a flip in the quantum state—that occurs whenever the fermion is swapped with itself. This sign change is unavoidable and permanent; it cannot be erased by simply redefining the particles or changing the perspective.

The researchers tested their findings against three well-known examples of these quantum systems to ensure the theory held up in concrete cases. In one example, known as the Laughlin state, which describes a specific type of electron fluid, the fermion attachment shifts the electric charge of every vortex by exactly one electron. In another example, involving a theory called SO(3)3, the attachment swaps two specific types of vortices while leaving a third one alone. In a third example, the Ising theory, which describes a system of free fermions, the attachment does not move any particles at all but instead flips the sign of the internal state of the vortices. In every single case, the result was the same: the symmetry could be lifted in exactly two ways, and these two ways were separated by the presence or absence of the fermion attached to the vortices.

This work is significant because it closes a long-standing gap in our understanding of how symmetries behave in these complex quantum environments. Before this proof, it was possible that there might be hidden, more complicated ways to lift these symmetries, or that the number of ways might depend on the specific details of the material. The new proof shows that the situation is remarkably simple and universal. The only freedom a physicist has when lifting a symmetry is this single binary choice. This clarity is essential for understanding the deeper properties of these materials, such as how they might store information or how they respond to external forces. By establishing that the ambiguity is strictly limited to this one specific operation, the researchers have provided a solid foundation for future studies of topological phases of matter, ensuring that the rules governing these exotic quantum worlds are as precise and predictable as the mathematics that describes them.

Technical Summary: Lifts of Symmetries to Minimal Nondegenerate Extensions

Problem Statement
The paper addresses the problem of lifting global symmetries from a (2+1)-dimensional fermionic topological order to its bosonic counterpart obtained by gauging fermion parity. Mathematically, a fermionic topological order is described by a slightly degenerate braided fusion category $\mathcal{C}$ (a super-modular category), characterized by a transparent fermion $\psi$ . Gauging fermion parity yields a minimal nondegenerate extension $\mathcal{E}$ (a modular category) containing $\mathcal{C}$ as a subcategory, where $\psi$ is no longer transparent.

The central question concerns the group of locality-respecting (LR) braided autoequivalences, denoted $\text{Aut}_{LR}(\mathcal{C})$ and $\text{Aut}_{LR}(\mathcal{E})$ . These groups classify symmetries that act trivially on the local fermion's fusion space and respect the fermion marking. The authors investigate the restriction homomorphism $r: \text{Aut}_{LR}(\mathcal{E}) \to \text{Aut}_{LR}(\mathcal{C})$ . Specifically, they aim to determine the kernel of this map and prove its surjectivity. Previous work had conjectured that the kernel is exactly $\mathbb{Z}/2$ , generated by "fermion attachment" (attaching $\psi$ to every vortex), but a general proof for arbitrary slightly degenerate categories was lacking.

Methodology
The authors employ a direct tensor-categorical approach, avoiding reliance on unitarity or specific ribbon structures, though these are recovered in the unitary case. The proof strategy involves three main technical components:

Normalization and Module Classification:
Elements in the kernel of $r$ are represented by pairs $(\Phi, \text{id}_\psi)$ where $\Phi$ restricts to the identity on $\mathcal{C}$ . The category $\mathcal{E}$ admits a $\mathbb{Z}/2$ -grading $\mathcal{E} = \mathcal{C} \oplus \mathcal{V}$ , where $\mathcal{V}$ is the vortex sector. Using graded Morita theory, the authors show that the restriction of such a $\Phi$ to the vortex sector $\mathcal{V}$ , viewed as a left $\mathcal{C}$ -module functor, must be isomorphic to a right multiplication $R_a = -\otimes a$ by an object $a$ in the Müger center of $\mathcal{C}$ . Since the center is generated by $\psi$ , there are only two possibilities: $a \cong \mathbf{1}$ (identity action) or $a \cong \psi$ (fermion attachment).
Relative Rigidity:
To distinguish between different representatives within the same module type, the authors prove a relative rigidity lemma. They show that if a braided autoequivalence acts as the identity on $\mathcal{C}$ and as the identity module functor on $\mathcal{V}$ , then it is monoidally isomorphic to the identity functor on $\mathcal{E}$ via an isomorphism that is the identity on $\mathcal{C}$ . This implies that the "freedom" in the vortex sector is reduced to a single scalar phase, which can be removed by rephasing the vortices, provided the module type is trivial.
Fermion Attachment and Surjectivity:
The authors explicitly construct the fermion attachment functor $\alpha$ , which acts as the identity on $\mathcal{C}$ and maps $X \mapsto X \otimes \psi$ for $X \in \mathcal{V}$ . They prove that $\alpha$ is a non-trivial element of the kernel because its tensor structure involves the fermionic self-braiding $c_{\psi,\psi} = -\text{id}$ , preventing it from being isomorphic to the identity via a morphism that fixes $\psi$ .
For surjectivity, they utilize graph algebras. Given a symmetry $F$ of $\mathcal{C}$ , they construct a graph algebra in $\mathcal{E}^{\text{rev}} \boxtimes \mathcal{E}$ , condense it to a local module category (identified as the toric code), and then condense a boson to form a Lagrangian algebra. This algebra is shown to be the graph of a lift $H$ of $F$ to $\mathcal{E}$ .

Key Contributions and Results
The paper establishes the Main Theorem, which states that for a slightly degenerate braided fusion category $\mathcal{C}$ and a fixed minimal nondegenerate extension $\mathcal{E}$ , the restriction homomorphism fits into a central short exact sequence:
$1 \longrightarrow \mathbb{Z}/2 \xrightarrow{[\alpha_\psi]} \text{Aut}_{LR}(\mathcal{E}) \xrightarrow{r} \text{Aut}_{LR}(\mathcal{C}) \longrightarrow 1$
where $[\alpha_\psi]$ is the class of fermion attachment.

Key specific results include:

Surjectivity: Every symmetry class in $\mathcal{C}$ lifts to $\mathcal{E}$ .
Kernel Structure: The kernel is exactly $\mathbb{Z}/2$ , generated by fermion attachment. There are no other hidden symmetries in the kernel.
Uniqueness up to Fermion Attachment: Every symmetry class in $\mathcal{C}$ has exactly two lifts to $\mathcal{E}$ . These two lifts differ precisely by attaching the fermion $\psi$ to every vortex.
Distinguishing Feature: The two lifts are distinguished by the fermionic self-braiding sign ( $c_{\psi,\psi} = -\text{id}$ ) appearing in their tensor structure, even if they induce the same permutation of simple object labels.
Generalization: The proof holds for non-unitary categories and does not require a ribbon structure, relying only on the braided fusion category axioms.

Significance and Claims
The authors claim to provide the first direct tensor-categorical proof of the conjecture that the kernel of the restriction map is $\mathbb{Z}/2$ for all slightly degenerate braided fusion categories, extending previous results that were limited to super-Abelian cases or relied on higher-categorical machinery.

The paper emphasizes that this result clarifies the structure of symmetry lifting in fermionic topological phases. Physically, it confirms that the only ambiguity in lifting a symmetry to the gauged theory is the choice of whether to bind the local fermion to the vortices. This has direct implications for the computation of 't Hooft anomalies in fermionic systems, as the obstruction to lifting a group action to the bosonic theory is governed by this $\mathbb{Z}/2$ kernel. The authors also relate their findings to the theory of graded extensions and braided Picard 2-groups, identifying the kernel with the automorphism group of the extension in the groupoid of minimal nondegenerate extensions.

The results are illustrated through three physical examples: the Laughlin state ( $U(1)_k$ ), the $SO(3)_3$ theory, and the Ising topological order, demonstrating how the abstract tensor-categorical machinery manifests in concrete anyon models.

Technical Summary: Lifts of Symmetries to Minimal Nondegenerate Extensions

More like this