🔢 mathematics

The bulk one-arm exponent for the CLE $_{\kappa'}$ percolations

This paper determines the exact bulk one-arm exponent for CLE $_{\kappa'}$ percolations in the non-simple regime by leveraging the coupling between Liouville quantum gravity and SLE curves, thereby completing the set of arm exponents for the fuzzy Potts model and providing the first prediction for the bichromatic one-arm exponent of the critical 3-state Potts model.

Original authors: Haoyu Liu, Xin Sun, Pu Yu, Zijie Zhuang

Published 2026-07-17

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Haoyu Liu, Xin Sun, Pu Yu, Zijie Zhuang

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

Imagine the universe as a giant, chaotic dance floor where particles are constantly bumping into each other, forming clusters, breaking apart, and creating intricate patterns. In the world of physics, there's a special kind of dance called "percolation." It's like watching a drop of ink spread through a sponge, or a fire jumping from tree to tree in a forest. Scientists love to study these patterns because they reveal hidden rules about how nature organizes itself, from the flow of electricity in a material to the spread of a disease.

But here's the tricky part: when you zoom in really close, these patterns look like messy, jagged scribbles. When you zoom out, they look like smooth, flowing loops. To understand the "messy" stuff, physicists use a magical tool called the Conformal Loop Ensemble (CLE). Think of CLE as a collection of invisible, rubber-band loops floating in a fluid. These loops are special because they look the same no matter how much you stretch or twist the space they're in (a property called conformal invariance). They are the "perfect" shapes that nature creates when things are at a critical tipping point—like water just about to boil or a magnet just about to lose its magnetism.

Now, imagine you have a game where you color these rubber-band loops either red or blue based on a coin flip. This creates a "fuzzy" version of the loops, where big clusters of red or blue form. Physicists have been trying to figure out exactly how big these clusters can get. Specifically, they wanted to know the odds of a single, continuous path of one color stretching from the center of the universe all the way to the edge. This is called the "one-arm exponent." It's like asking, "What are the chances that a single red thread can weave its way from the middle of a tangled ball of yarn to the outside?" For a long time, scientists could calculate the odds for almost every scenario, except for this specific "center-to-edge" thread in the most complex version of the game.

The Paper's Discovery: Cracking the Code of the "Fuzzy" Loops

This paper by Haoyu Liu, Xin Sun, Pu Yu, and Zijie Zhuang is the story of finally solving that missing puzzle. The authors tackle the "bulk one-arm exponent" for a specific, complicated version of these rubber-band loops known as CLE percolations (specifically in the "non-simple" regime where loops can touch but not cross).

To understand what they did, imagine the loops as a set of Russian nesting dolls. The authors realized that to find the size of the big red or blue clusters, they didn't need to look at the whole messy ball of yarn at once. Instead, they could build the picture piece by piece, starting from the boundary and working inward. They used a technique called conformal welding, which is like taking two pieces of stretchy, quantum fabric and sewing them together perfectly along their edges. If the edges match up just right, the fabric snaps into a new, stable shape.

The authors used a very advanced mathematical framework called Liouville Quantum Gravity (LQG). You can think of LQG as a way to describe a surface that is constantly rippling and changing shape, like a trampoline made of liquid. By "welding" these quantum surfaces together, the authors could translate the messy problem of the loops into a clean problem about the geometry of these surfaces. They discovered that the answer to the "one-arm" question is hidden in the "structure constants" of a theory called Liouville Conformal Field Theory. It's as if they found a secret code in the fabric of space-time that tells you exactly how likely a red thread is to reach the edge.

The Big Reveal

The paper proves that this missing exponent exists and gives a precise formula for it. The formula is the unique positive solution to a specific equation involving sine functions and the parameters of the model (called $\kappa'$ and $\rho$ ).

Why does this matter? Because this exponent tells us the fractal dimension of the clusters. In simple terms, it tells us how "space-filling" the red or blue clusters are. A dimension of 2 means they fill the whole space; a dimension of 1 means they are just thin lines. The authors show that for the critical 3-state Potts model (a famous model in physics describing how magnets with three possible states behave), the "bichromatic" one-arm exponent (where a path can be made of two different colors) is exactly 4/135.

This is a big deal because, as the authors note, this specific number 4/135 had never been predicted by physicists before. It was a mystery. By solving the math for the continuous loops, they have provided the exact answer for the discrete 3-state Potts model, assuming the standard belief that the discrete model converges to the continuous loop model.

What They Didn't Do (and What They Did)

It is important to note what this paper is not. They did not simulate the model on a computer to guess the answer. They did not suggest that the answer is "probably" 4/135. They proved it mathematically. They derived an exact formula based on the rigorous properties of the loops and the quantum surfaces.

They also didn't just make up a new theory; they built upon existing, well-established ideas like the FK conformal invariance conjecture (the idea that the discrete FK percolation models turn into these continuous loops when you zoom out). They used this conjecture to bridge the gap between their continuous math and the discrete physics models.

The Takeaway

In the end, this paper is a triumph of pure mathematical reasoning. The authors took a problem that had stumped researchers for years—the size of the "arm" in a fuzzy, multi-colored loop system—and solved it by weaving together ideas from probability, geometry, and quantum field theory. They showed that the universe's "fuzzy" patterns follow a strict, solvable rule, and for the first time, we know exactly what that rule is for the 3-state Potts model. The number 4/135 is no longer a guess; it is a fact derived from the deep structure of the mathematical universe.

Problem Statement
The paper addresses the calculation of the bulk one-arm exponent for Conformal Loop Ensemble (CLE) percolations in the non-simple regime $\kappa' \in (4, 8)$ . CLE $_{\kappa'}$ describes the scaling limit of critical Fortuin–Kasteleyn (FK) percolations. CLE percolations, introduced by Miller, Sheffield, and Werner (MSW), describe the scaling limit of the "fuzzy Potts model," a variant of FK percolation with an additional coloring parameter $r$ . While K"ohler-Schindler and Lehmkuehler (KSL) previously derived all bulk and boundary arm exponents for the fuzzy Potts model under the assumption of FK conformal invariance, the bulk one-arm exponent remained unsolved. This exponent is crucial as it prescribes the fractal dimension of the clusters in CLE $_{\kappa'}$ percolations. Specifically, the paper aims to derive the exact value of this exponent in terms of the parameters $\kappa'$ and $\rho$ (where $\rho$ relates to the coloring probability $r$ ), and subsequently apply this to the discrete fuzzy Potts model and the critical 3-state Potts model.

Methodology
The authors employ a rigorous approach based on the coupling between Liouville Quantum Gravity (LQG) and Schramm-Loewner Evolution (SLE) curves, a framework known as the "mating-of-trees" theory. The derivation proceeds through the following steps:

Iterative Construction via BCLE: The authors utilize the iterative construction of CLE percolations from the Boundary Conformal Loop Ensemble (BCLE) established in MSW and KSL. The bulk one-arm exponent is encoded in the distribution of the conformal radius of loops surrounding the origin in the BCLE.
Conformal Welding of Quantum Surfaces: The core technical innovation is the use of conformal welding to characterize BCLE loops.
- For the simple regime ( $\kappa \in (2, 4)$ , where $\kappa = 16/\kappa'$ ), the authors prove that a BCLE loop can be realized as the interface resulting from the uniform conformal welding of a quantum disk and a "pinched thin quantum annulus."
- For the non-simple regime ( $\kappa' \in (4, 8)$ ), they extend this to generalized quantum surfaces (specifically "forested" quantum surfaces) and prove that the BCLE loop arises from the welding of a forested quantum disk and a forested pinched thin quantum annulus.
Liouville Conformal Field Theory (LCFT): To extract the exact moments of the conformal radius from these welding results, the authors leverage the exact solvability of boundary Liouville Conformal Field Theory. They utilize structure constants and boundary Liouville reflection coefficients (derived in works by RZ) to compute the necessary integrals.
Target-Invariant SLE: A key technical step involves proving conformal welding results for target-invariant radial SLE curves, which allows the characterization of the BCLE loop surrounding the origin.

Key Contributions and Results

Exact Solution for the Bulk One-Arm Exponent: The paper derives an exact formula for the bulk one-arm exponent $\alpha_1(r)$ of CLE $_{\kappa'}$ percolations. The exponent is defined as the unique positive solution to a specific trigonometric equation involving $\kappa$ and $\rho$ (Theorem 1.1).
Extension to the Fuzzy Potts Model: Assuming the FK conformal invariance conjecture, the authors show that the bulk one-arm exponent for the discrete fuzzy Potts model is identical to that of the continuum CLE percolation (Theorem 1.2).
New Results for BCLE: As intermediate results, the paper provides exact formulae for the probability that the origin is surrounded by a clockwise (true) or counterclockwise (false) loop in BCLE, as well as the moments of the conformal radius of the domain enclosed by such loops (Theorems 1.5–1.8). These results hold for both simple ( $\kappa \in (2, 4)$ ) and non-simple ( $\kappa' \in (4, 8)$ ) regimes, as well as the critical case $\kappa=4$ .
Application to the 3-State Potts Model: As a specific corollary, the authors calculate the bichromatic one-arm exponent for the critical 3-state Potts model. By setting $q=3$ and $r=1/3$ , they derive that this exponent is exactly $4/135$ (Corollary 1.4).

Significance and Claims
The paper claims that the derived bulk one-arm exponent for the fuzzy Potts model was not previously predicted in physics literature. The authors emphasize that their derivation relies on the structure constants of boundary Liouville conformal field theory, providing a new method to solve for exponents that were previously inaccessible.

The work extends the understanding of the fractal dimensions of clusters in CLE percolations and the fuzzy Potts model. Specifically, it resolves the missing exponent that determines the dimension of the clusters, complementing previous work on the CLE gasket and carpet dimensions. The result for the 3-state Potts model ( $4/135$ ) is presented as a novel prediction for a natural exponent in a model of substantial interest to the physics community.

The authors maintain a modest tone regarding the scope, noting that their results for the fuzzy Potts model depend on the standard FK conformal invariance conjecture. They also note that while their method relies on the integrability of LCFT, it remains an open question whether the bulk one-arm exponent can be derived without this integrability, perhaps via techniques involving L'evy processes as seen in related MSW works.

More like this