On the action of non-invertible symmetries on local operators in 3+1d

This paper establishes that finite non-invertible symmetries in 3+1d without topological line operators necessarily act invertibly on local operators, leading to a decomposition of general non-invertible symmetries into invertible actions and gauging interfaces, and deriving necessary conditions for such symmetries to be anomaly-free.

The Gist

Imagine the universe as a giant, complex dance floor. In physics, the "dancers" are particles and fields, and the "rules of the dance" are symmetries. For a long time, physicists thought these rules were like a standard dance troupe: if you did a move, you could always do the exact opposite move to get back to where you started. This is called an invertible symmetry.

But recently, physicists discovered a new, weird kind of dance rule called non-invertible symmetry. Imagine a move where, if you do it, you can't just "undo" it with a single step. It's like a magic trick where you turn a red ball into a blue one, but there's no single button to turn the blue ball back into a red one.

This paper asks a very specific question: How do these weird, non-reversible magic tricks affect the individual dancers (local operators) on the floor?

Here is the breakdown of their discovery, using simple analogies:

1. The "Ghost" vs. The "Real" Move

The authors realized that in our 4-dimensional world (3 dimensions of space + 1 of time), these non-invertible symmetries come in two flavors:

• The "Ghost" Symmetries: Some of these symmetries are like ghosts. They exist as large, floating membranes in the universe, but they don't actually touch or change the individual dancers. If you try to use a "ghost" symmetry on a single particle, nothing happens. The paper proves that if a symmetry has no "lines" (thin, string-like defects) floating around, it must act like a normal, reversible symmetry on the particles. It's like a ghost trying to push a car; it looks scary, but it can't actually move the car.
• The "Real" Symmetries: The only time these symmetries actually scramble the particles in a non-reversible way is when they are built from a specific recipe: Gauging.

2. The "Gauging" Recipe (The Coset)

The paper explains that the only way to get a truly non-invertible effect on particles is to take a normal group of dancers (an invertible symmetry) and perform a "Gauging" operation.

The Analogy:
Imagine a dance troupe where everyone is wearing a specific colored shirt (a symmetry).

1. Normal Symmetry: You can swap the shirts around, but everyone is still distinct.
2. Gauging: You decide to "erase" the distinction between everyone wearing a red shirt. You force them to act as a single unit.
3. The Result: Now, if you try to look at a single dancer, you can't tell if they were originally red or blue because the "erasing" process has mixed them up. The symmetry is now non-invertible.

The authors show that every non-invertible symmetry acting on particles is just a combination of:

1. A normal, reversible dance move (the invertible part).
2. Followed by this "erasing/mixing" gauging process.

So, the "magic" isn't actually new magic; it's just a standard dance move followed by a specific type of cleanup crew that mixes things up.

3. The "Anomaly" Check (Is the Dance Floor Safe?)

In physics, an "anomaly" is like a glitch in the matrix. It means the rules of the dance are so contradictory that the universe can't exist in a stable, empty state (a "trivially gapped phase").

The paper asks: When is this non-invertible dance safe? When does it not cause a glitch?

They found a strict condition for safety. For a non-invertible symmetry to be "anomaly-free" (safe), the groups involved in the "Gauging" recipe must fit together perfectly, like two puzzle pieces that interlock without overlapping or leaving gaps.

Mathematically, they call this a Zappa-Szép product (or bicrossed product).

• Simple Analogy: Imagine you have a group of people (Group H) and another group (Group K). To make a safe, non-invertible symmetry, these two groups must be able to combine to form a bigger group (Group G) in a very specific way where they don't step on each other's toes, but they also don't leave any empty space. If they don't fit this specific "lock-and-key" pattern, the universe would glitch out.

4. The Big Takeaway

The paper concludes that in our 4D universe, non-invertible symmetries acting on particles are not "intrinsically" weird.

• If they don't have "lines" attached to them, they act exactly like normal, reversible symmetries.
• If they do act weirdly, it's because they are just a normal symmetry plus a "gauging" interface (a mixing process).

Why does this matter?
It simplifies the universe. Instead of thinking there are thousands of brand-new, mysterious laws of physics, the authors show that these new laws are actually just old laws wearing a disguise. They are built from familiar ingredients (groups and gauging) arranged in a specific way.

Summary in One Sentence

The paper proves that in our 4D world, if a "non-reversible" symmetry tries to change a particle, it's either doing nothing at all, or it's just a normal reversible move followed by a specific "mixing" process, and for the universe to remain stable, the groups involved in that mixing must fit together like perfect puzzle pieces.

Technical Summary

Here is a detailed technical summary of the paper "On the action of non-invertible symmetries on local operators in 3+1d" by Pavel Putrov and Rajath Radhakrishnan.

1. Problem Statement

The paper addresses a fundamental question in the classification of generalized symmetries in Quantum Field Theory (QFT): How do general finite non-invertible symmetries act on local operators in spacetime dimensions $d > 2$ (specifically 3+1d)?

While it is known that in 2+1 dimensions, non-invertible symmetries (implemented by topological surface operators) can act non-invertibly on local operators, the situation in higher dimensions is less clear. Previous examples of non-invertible symmetries in 3+1d include:

• Condensation operators: Act trivially on local operators.
• Duality defects: Act invertibly (via gauging and relabeling).
• Coset symmetries: Can act non-invertibly, but their action is reducible to an invertible group action combined with a gauging interface.

The authors investigate whether all non-invertible symmetries in 3+1d fall into these categories or if there exist "intrinsically" non-invertible actions on local operators that cannot be decomposed into simpler invertible actions and gauging interfaces.

2. Methodology

The authors employ a combination of topological quantum field theory (TQFT) tools, fusion category theory, and dimensional reduction arguments.

• Equivalence Classes of Operators: They establish that the action of a topological codimension-1 operator (membrane in 3+1d, surface in 2+1d) on a local operator is invariant if the operators are connected by a topological interface of codimension-2. Thus, the analysis focuses on the equivalence classes of these operators (Schur components), denoted as $\pi_0(\mathcal{C})$.
• Symmetry Topological Field Theory (SymTFT): They utilize the $(d+1)$-dimensional SymTFT $Z(\mathcal{C})$ associated with a $d$-dimensional symmetry category $\mathcal{C}$. By analyzing the boundary conditions and the bulk-to-boundary maps, they relate the properties of the symmetry in $d$ dimensions to the structure of the bulk TQFT.
• Dimensional Reduction: They analyze the behavior of topological membrane operators (in 3+1d) and surface operators (in 2+1d) by compactifying them on spheres ($S^{d-1}$) to produce lower-dimensional line operators.
• Gauging Interfaces: They utilize the concept of gauging subgroups (or higher-form symmetries) to construct interfaces between the original symmetry category and a "gauged" category where line operators are removed.

3. Key Contributions and Results

A. Action on Local Operators in 2+1d (Review and Extension)

• No Line Operators: If a fusion 2-category $\mathcal{C}$ describing a 2+1d symmetry contains no non-trivial topological line operators ($\Omega\mathcal{C} \cong \text{Vec}$), the authors prove that all surface operators in $\mathcal{C}$ are invertible. Consequently, they act invertibly on local operators. This is shown via the 3+1d SymTFT (which becomes a Dijkgraaf-Witten theory) and dimensional reduction arguments.
• General Case: For general non-invertible symmetries containing line operators, the action on local operators can be decomposed. Specifically, any surface operator $S$ acts as:
  $$S(O) = I^\dagger \left( \sum n_i P_i(O) \right)$$
  where $P_i$ are invertible surface operators in a gauged theory, and $I^\dagger$ is the action of a gauging interface. This confirms that non-invertible actions arise from the interplay of invertible symmetries and gauging.

B. Action on Local Operators in 3+1d (Main Novelty)

• The "No Line" Theorem: The authors prove a central result: If a finite non-invertible symmetry in 3+1d has no topological line operators ($\Omega^2\mathcal{C} \cong \text{Vec}$), it acts invertibly on local operators.
  • Mechanism: They show that the set of equivalence classes of membrane operators $\pi_0(\mathcal{C})$ admits a group-like fusion rule isomorphic to a finite group $G$. The local operators transform in a representation of this group $G$.
  • Proof: Using the 4+1d SymTFT, they demonstrate that gauging the line operators (which are trivial) leads to a theory where membrane operators form a group extension. Dimensional reduction of membranes on $S^2 \times \mathbb{R}$ further confirms that simple membranes must be invertible.
• General Decomposition: For general non-invertible symmetries with topological line operators ($\Omega^2\mathcal{C} \cong \text{Rep}(H)$), the action on local operators decomposes into:
  1. An invertible action by membrane operators in a "line-free" theory (obtained by gauging the $H$ 2-form symmetry).
  2. The action of a gauging interface $I^\dagger$ associated with gauging the $H$ symmetry.
     This generalizes the behavior of coset symmetries to all finite non-invertible symmetries in 3+1d.

C. Anomaly-Free Non-Invertible Symmetries

The paper derives necessary conditions for a non-invertible symmetry to be anomaly-free (i.e., admitting a trivially gapped phase).

• Without Line Operators: An anomaly-free non-invertible symmetry without line operators is non-intrinsically non-invertible. It can be rendered invertible (up to condensation defects) via topological manipulation.
• With Line Operators (General Case): For a symmetry described by $\mathcal{C}$ with $\Omega^2\mathcal{C} \cong \text{Rep}(H)$ to be anomaly-free, the bulk SymTFT $Z(\mathcal{C})$ must have a line structure $\Omega^2 Z(\mathcal{C}) \cong \text{Rep}(G)$. The authors derive a necessary group-theoretic condition:
  $$G \cong H \rtimes K$$
  where $K$ is a subgroup of $G$ such that $G$ is a Zappa-Szép product (or bicrossed product) of $H$ and $K$. This implies that the double cosets $H \backslash G / H$ label the membrane operators, and the symmetry structure is tightly constrained by the group theory of $G, H,$ and $K$.

4. Significance

• Classification of Symmetries: The paper significantly narrows the landscape of possible non-invertible symmetries in 3+1d. It establishes that "genuine" non-invertible actions on local operators are rare; they essentially always reduce to invertible group actions modified by gauging interfaces.
• Intrinsic vs. Non-Intrinsic: It provides a rigorous criterion for "intrinsic" non-invertibility. In 3+1d, if a symmetry has no line operators, it is never intrinsically non-invertible regarding its action on local operators.
• Anomaly Constraints: The derived condition ($G = H \rtimes K$) for anomaly-free symmetries provides a powerful tool for constructing or ruling out candidate non-invertible symmetries in physical models.
• Generalizability: The arguments are constructed to generalize to higher dimensions ($d+1$) and to cases involving fermionic line operators (SVec), suggesting a universal structure for non-invertible symmetries in dimensions $d \geq 2$.

5. Conclusion

The authors conclude that in 3+1d, the action of any finite non-invertible symmetry on local operators is effectively "invertible" up to the insertion of a gauging interface. This implies that the rich structure of non-invertible symmetries in higher dimensions is largely governed by the interplay between invertible 0-form symmetries and the gauging of higher-form symmetries, rather than by fundamentally new types of non-invertible algebraic structures acting directly on local fields. The paper leaves open the question of whether anomaly-free symmetries exist that cannot be constructed from finite groups, but provides strong evidence that such cases are highly constrained.