Table of Contents
Fetching ...

Asymptotic density of states in 2d CFTs with non-invertible symmetries

Ying-Hsuan Lin, Masaki Okada, Sahand Seifnashri, Yuji Tachikawa

TL;DR

This work generalizes Cardy’s universal density of states to 2d CFTs with finite, non-invertible symmetries described by a fusion category $\mathcal{C}$. It develops two complementary frameworks—the 2d tube-algebra view and the 3d bulk TV$_{\mathcal{C}}$ perspective with Drinfeld center $\mathcal{Z}(\mathcal{C})$—to classify irreducible sectors as simple objects of $\mathcal{Z}(\mathcal{C})$ and to derive their asymptotic counts. The main result states that the density in the $a$-twisted sector transforming in the center-label $\mu$ scales as $\langle \mu,a\rangle\dim\mu$, up to the universal Cardy growth $e^{(L/T)\pi c/6}$; in the finite-group limit this reproduces Pal–Sun’s $(\dim\rho)^2$ formula and extends naturally to modular and Haagerup fusion-category symmetries. The findings provide a unified description of state densities and selection rules across a broad class of fusion-category symmetries, with concrete illustrations for finite groups, diagonal RCFTs, and the Haagerup case.

Abstract

It is known that the asymptotic density of states of a 2d CFT in an irreducible representation $ρ$ of a finite symmetry group $G$ is proportional to $(\dimρ)^2$. We show how this statement can be generalized when the symmetry can be non-invertible and is described by a fusion category $\mathcal{C}$. Along the way, we explain what plays the role of a representation of a group in the case of a fusion category symmetry; the answer to this question is already available in the broader mathematical physics literature but not yet widely known in hep-th. This understanding immediately implies a selection rule on the correlation functions, and also allows us to derive the asymptotic density.

Asymptotic density of states in 2d CFTs with non-invertible symmetries

TL;DR

This work generalizes Cardy’s universal density of states to 2d CFTs with finite, non-invertible symmetries described by a fusion category . It develops two complementary frameworks—the 2d tube-algebra view and the 3d bulk TV perspective with Drinfeld center —to classify irreducible sectors as simple objects of and to derive their asymptotic counts. The main result states that the density in the -twisted sector transforming in the center-label scales as , up to the universal Cardy growth ; in the finite-group limit this reproduces Pal–Sun’s formula and extends naturally to modular and Haagerup fusion-category symmetries. The findings provide a unified description of state densities and selection rules across a broad class of fusion-category symmetries, with concrete illustrations for finite groups, diagonal RCFTs, and the Haagerup case.

Abstract

It is known that the asymptotic density of states of a 2d CFT in an irreducible representation of a finite symmetry group is proportional to . We show how this statement can be generalized when the symmetry can be non-invertible and is described by a fusion category . Along the way, we explain what plays the role of a representation of a group in the case of a fusion category symmetry; the answer to this question is already available in the broader mathematical physics literature but not yet widely known in hep-th. This understanding immediately implies a selection rule on the correlation functions, and also allows us to derive the asymptotic density.
Paper Structure (29 sections, 137 equations, 5 figures)

This paper contains 29 sections, 137 equations, 5 figures.

Figures (5)

  • Figure 1: States in the sector $V_c\otimes\overline{V_d}\subset\mathcal{H}_a$ can be realized in the 3d TQFT description as in this figure. Associated with each end $T^2$ are holomorphic and antiholomorphic boundary conditions. Take operators $\mathcal{O}_c$ and $\overline{\mathcal{O}_d}$ on the boundaries, corresponding to states in $V_c$ and $\overline{V_d}$, respectively. They are connected by topological lines labeled by $c$ and $d$ meeting in the middle. Choose $x_{cd}^a\in\mathrm{Hom}(a\otimes d,c)$ to fuse $c$ and $d$ into $a$. Then $\mathcal{O}_{cd}^{a;x}=(\mathcal{O}_c,\mathcal{O}_d,x_{cd}^a)$ is the defect operator corresponding to a state in the sector $V_c\otimes\overline{V_d}\subset\mathcal{H}_a$ (See footnote \ref{['folding']}). The degree of freedom in choosing $x_{cd}^a$ is $\dim \mathrm{Hom}(a\otimes d,c)=N_{ad}^c$, which leads to the multiplicity of the sector $V_c\otimes\overline{V_d}$ in $\mathcal{H}_a$.
  • Figure 2: The 2d theory $Q$ on $\Sigma$ is obtained by putting the 3d TQFT $\mathrm{TV}_\mathcal{C}$ on $\Sigma \times I$. We have the non-topological boundary condition $\mathbb{B}_Q$ on the left and the (Dirichlet) topological boundary condition $\mathbb{D}$ on the right. The topological lines $a,b,c,\dots \in \mathcal{C}$ on the $\mathbb{D}$ boundary implement the $\mathcal{C}$ symmetry of $Q$.
  • Figure 3: A defect operator $\mathcal{O}\in\mathcal{H}_a$ of the 2d theory $Q$ is composed of a bulk anyon $\mu\in\mathcal{Z}(\mathcal{C})$ and a pair of boundary (defect) operators. Via the state-operator correspondence, $\widetilde{\mathcal{O}}$ belongs to the $\mu$-punctured disk Hilbert space of $\mathrm{TV}_\mathcal{C}$ with boundary condition $\mathbb{B}_Q$, and $x$ belongs to $\mathrm{Hom}_\mathcal{C}(F(\mu),a)$ where $F:\mathcal{Z}(\mathcal{C}) \to \mathcal{C}$ describes the fusion of bulk anyons with the topological boundary $\mathbb{D}$.
  • Figure 4: The partition function of $\mathrm{TV}_\mathcal{C}$ on the solid torus $D_L^2 \times S_T^1$ with $\mathbb{B}_Q$ boundary condition and anyon $\mu$ in the middle. $D_L^2$ is the horizontal disk of circumference $L$, and $S^1_T$ is the vertical circle of size $T$. Since the boundary condition $\mathbb{B}_Q$ is conformal, this partition function depends only on $q=e^{-2\pi(T/L)}$.
  • Figure 5: Boundary crossing relation in 2+1d TQFT. The blue dashed lines sketch the surfaces of the gapped boundaries, such that the regions outside the slabs and inside the cylinders are empty. Top: A special case that involves the half-linking matrices $\Psi_{a(\mu,xy)}$. Bottom: The general case.