⚛️ high-energy theory

Conformal correlator systems

This paper introduces "conformal correlator systems" as a framework with weaker axioms than conformal field theory to describe non-local, conformally covariant correlators in critical statistical models, demonstrating that such systems are determined by four-point functions and can exhibit arbitrary complex conformal spins with nontrivial monodromies.

Original authors: Sylvain Ribault

Published 2026-09-28

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Sylvain Ribault

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 world of statistical physics, scientists often study how tiny particles, like spins on a grid, behave when a material is pushed to a critical point. At this precise moment of transition, the system becomes scale-invariant, meaning it looks the same whether viewed from a mile away or a millimeter away. For decades, the standard tool for describing these critical systems has been conformal field theory. This framework relies on a powerful idea called the operator product expansion, which essentially states that when two fields get very close together, they can be replaced by a simple sum of other fields. This rule acts like a local dictionary, allowing physicists to break down complex interactions into manageable pieces. However, this dictionary has a limitation: it assumes that the variables describing the system are local, existing at specific points in space. But in many modern models, particularly those involving clusters or loops that stretch across the entire system, the variables are non-local. They do not live at a single point but connect distant regions, and for these systems, the standard dictionary fails. The question arises: if these systems still possess the beautiful symmetry of scale invariance, but refuse to follow the local rules of the standard theory, what mathematical structure do they actually belong to?

Sylvain Ribault, a theoretical physicist at the Institut de physique théorique in France, proposes a new framework to answer this question, one that he calls a conformal correlator system. Instead of forcing these non-local systems into the rigid box of traditional conformal field theory, Ribault relaxes the rules just enough to let them breathe. In a standard theory, the behavior of a system with many points is determined entirely by the interactions of just three points. Ribault shows that in these generalized systems, the behavior of the whole is determined by the interactions of four points. This shift from three to four is not a minor adjustment; it reflects a fundamental change in how information is shared across the system. Because the variables are non-local, the interaction between two points cannot be described in isolation; it depends on the entire context of the system, much like how the meaning of a word in a sentence depends on the surrounding words rather than just the two words next to it. By defining a set of axioms that allow for these context-dependent interactions, Ribault creates a consistent mathematical home for systems that were previously difficult to categorize.

The paper explores this new framework primarily in two dimensions, where the mathematics of symmetry is richest. In this setting, the theory allows for a surprising freedom: the "spins" of the fields, which usually must be whole numbers or simple fractions, can now take on any complex value. This freedom leads to a phenomenon where the system's properties change in a specific, predictable way when one point is moved around another. In the standard theory, moving a point around another leaves the system unchanged. In Ribault's generalized systems, the system picks up a phase factor, a kind of mathematical twist, which can be simple or complex. Ribault demonstrates that these twists are not random; they are constrained by strict rules. For instance, on a sphere, the twists associated with four points must satisfy a specific balance equation. Similarly, on a torus, which is a shape like a donut, the properties of a single field are constrained by how the system behaves when the shape is stretched or twisted. These constraints act as a filter, allowing only certain configurations of fields to exist, and Ribault shows that these filters are consistent with the new, weaker rules of the theory.

To prove that this abstract framework is not just a mathematical curiosity but a tool for real physics, Ribault applies it to several known models. He looks at critical loop models, which describe the connectivity of clusters in systems like the Potts model, and free bosonic theories, which describe waves in a field. In the case of critical loop models, the theory successfully reproduces the known results for single-valued systems while also predicting new, multi-valued versions where the fields carry non-trivial twists. He also examines topological defects, which are lines or boundaries that separate different regions of a material. When a field moves around such a defect, the system's topology changes, and the standard rules break down. Ribault's framework handles this naturally, treating the defect as an extra parameter that modifies the interaction rules without breaking the underlying symmetry. The results show that the new system can describe these complex, non-local behaviors without losing the predictive power of conformal symmetry.

The paper also addresses the question of how to solve these systems. In traditional conformal field theory, physicists often solve for the three-point interactions and then build up to larger systems. Ribault shows that in his generalized systems, one must solve for the four-point interactions first. He provides numerical evidence that for certain models, the number of possible solutions matches the number of known physical configurations, suggesting that the framework is complete. He also notes that while the standard theory requires the system to be single-valued, these new systems can be multi-valued, opening the door to describing fields with arbitrary spins. This is a significant departure from the past, where such fields were often considered unphysical or too difficult to handle. By allowing these spins to vary, the theory simplifies the mathematical expressions for the interactions, replacing complicated trigonometric coefficients with simpler exponential forms.

Ultimately, this work offers a broader perspective on what a physical theory can be. It suggests that the rigid requirement of local interactions is not a fundamental law of nature but a convenient approximation that works for many, but not all, systems. By introducing the concept of a correlator expansion, where the interaction coefficients depend on the whole system rather than just the local pair, Ribault provides a language for describing the non-local reality of critical phenomena. The framework does not discard the old theory but rather extends it, showing that the beautiful symmetry of scale invariance can survive even when the local rules of interaction are relaxed. This opens up new avenues for understanding complex materials and statistical models, offering a way to describe systems that were previously thought to be beyond the reach of conformal symmetry. The work stands as a testament to the flexibility of mathematical physics, showing that by carefully adjusting the axioms, one can capture a wider range of the universe's behaviors.

Technical Summary: Conformal Correlator Systems

1. Problem Statement

In statistical physics, critical limits of models such as the $Q$ -state Potts model (generalized to non-integer $Q$ ) and critical loop models exhibit conformal symmetry. However, these systems often involve non-local variables (e.g., cluster connectivities) that do not admit Operator Product Expansions (OPEs). Standard Conformal Field Theory (CFT) relies fundamentally on the existence of OPEs, where the product of two local fields is replaced by a linear combination of fields with coefficients independent of the rest of the correlator.

The absence of OPEs in these non-local correlators creates a theoretical gap:

Standard CFT axioms fail because correlators cannot be reduced to 3-point structure constants via OPEs.
Despite the lack of OPEs, these correlators are observed to be linear combinations of conformal blocks, yet the coefficients do not factorize as expected in CFT.
Existing frameworks struggle to describe correlators with non-integer conformal spins and non-trivial monodromies (multivaluedness) without forcing them into an operator formalism that may not exist (e.g., the torus partition function).

The paper asks: What conformally invariant structure do these cluster connectivities belong to if not a CFT?

2. Methodology and Definitions

The author proposes a generalization of CFT by weakening the axioms while retaining the core structure of conformal covariance and block decomposition.

2.1 Core Definitions

Correlator Expansion (Axiom 2.6): A generalization of the OPE. For $N \ge 2$ , an $N$ -point correlator is expanded into $(N-1)$ -point correlators. Crucially, the expansion coefficients ( $C_{R}^{p,q}$ ) may depend on the entire correlator (including extra parameters $p$ and other fields), not just the two fields being expanded. This makes the expansion non-local.
Conformal Correlator System (Definition 2.9): A set of correlators that transform covariantly under the conformal algebra and admit correlator expansions (instead of OPEs).
2d Conformal Correlator System (Definition 3.6): Extends the definition to two dimensions by:
1. Allowing local conformal symmetry (Virasoro algebra).
2. Allowing fields to have arbitrary complex conformal spins $S \in \mathbb{C}$ .
3. Requiring "handle-cutting" (reduction of genus $g$ to $g-1$ ) but relaxing the constraints on the structure constants $H_R$ compared to standard CFT.

2.2 Key Modifications to Standard CFT

Spectrum: Spins $S$ are not restricted to integers or half-integers. This allows for multivalued correlators with non-trivial monodromies.
Torus Partition Function: The paper distinguishes between the operator formalism (trace over states, requiring modular invariance of the partition function) and the bootstrap approach. In conformal correlator systems, a torus 0-point correlator exists and is modular invariant, but it may sum over only a subset of the spectrum. Fields with non-integer spins may exist in the system even if they do not contribute to the torus 0-point correlator.
Monodromies: The paper introduces phases $\theta = e^{2\pi i S}$ $θ = e^{2 π i S}$ . Correlators are classified by their monodromy properties:
- Abelian: The correlator picks up a constant prefactor under braiding.
- Non-abelian: The correlator undergoes a matrix transformation.

3. Key Contributions and Results

3.1 Reduction to 4-Point Correlators

In standard CFT, $N$ -point correlators are determined by 3-point structure constants due to the factorization of OPEs. In conformal correlator systems, the author proves (Proposition 2.25) that all $N$ -point correlators are determined by 4-point correlators (up to constant factors).

Mechanism: Crossing symmetry equations relate different channel decompositions. Any change of basis for $N$ -point blocks can be decomposed into changes of basis for 4-point blocks.
Implication: The fundamental data required to solve the theory shifts from 3-point to 4-point structure constants.

3.2 Monodromy Constraints

The paper derives specific constraints on the phases of fields in correlators with abelian monodromies:

Sphere 4-point Correlators: For a correlator $\langle V_1 V_2 V_3 V_4 \rangle$ , the phases $\theta_s, \theta_t, \theta_u$ of the intermediate fields in the $s, t, u$ channels must satisfy:
$\theta_s \theta_t \theta_u = \theta_1 \theta_2 \theta_3 \theta_4$
(Equation 1.5). This implies a constraint on the sum of spins modulo integers.
Torus 1-point Correlators: For a correlator $\langle V_1 \rangle^{(1)}$ , the phase $\theta$ of the channel fields must satisfy:
$\theta^{12} = \theta_1$
(Equation 1.6). This implies that if a field has a non-trivial phase, the spectrum must contain fields with phases that are 12th roots of that phase, leading to an infinite tower of phases in the spectrum.

3.3 Modular Covariance

The paper generalizes modular covariance to include multiplier systems (Definition 3.25) to accommodate arbitrary spins. A torus 1-point correlator is modular covariant with weights $(\Delta, \bar{\Delta})$ and phase $\theta_1$ if it transforms with a multiplier system $\epsilon(g)$ . The relation $\epsilon(S^2)\sqrt{\theta_1} = 1$ is derived, linking the multiplier system to the spin.

3.4 Examples and Applications

The paper classifies and analyzes several systems within this framework:

Critical Loop Models: These are the primary motivation. They possess single-valued correlators with no OPEs. The paper proposes "deformed" (non-integer spins, abelian monodromies) and "extended" (non-abelian monodromies) generalizations.
- Result: Numerical solutions of crossing symmetry for 4-point correlators in deformed models show that the dimension of the solution space remains finite and matches the number of weakly connected combinatorial maps, provided the monodromy constraint (1.5) is satisfied.
Free Bosons: A system with OPEs but multivalued correlators (if left/right momenta are decoupled). Monodromies are abelian.
Coulomb Gas Integrals: Generalizations of free bosons. While standard integrals are single-valued, allowing $\alpha \neq \bar{\alpha}$ leads to non-abelian monodromies (e.g., in critical XXZ spin chains).
Topological Defects: A CFT with topological defects can be reinterpreted as a conformal correlator system where the defect acts as an extra parameter. The defect modifies the expansion coefficients, breaking the standard OPE factorization and potentially introducing non-integer spins.

4. Significance and Claims

The paper claims that conformal correlator systems provide a necessary and consistent framework for describing critical statistical models that possess conformal symmetry but lack OPEs.

Theoretical Necessity: The author argues that forcing these models into a standard CFT framework (by inventing non-local objects) is more complicated than directly studying the correlators' properties via the weaker axioms of correlator expansions.
Solvability: The framework suggests that solving these models requires determining 4-point structure constants rather than 3-point ones.
Simplification: Allowing arbitrary complex spins may simplify the structure constants. For instance, in critical loop models, trigonometric coefficients depending on Kac indices may become simpler when expressed as functions of $e^{i\pi s}$ .
Unification: The framework unifies various models (critical $O(n)$ , Potts, PSU( $n$ )) as subsystems of a larger conformal correlator system, valid for generic values of parameters where they are not standard CFTs.

The paper concludes that while the axiomatic approach is established, constructive methods (such as Schramm–Loewner Evolutions or Conformal Loop Ensembles) may be required to explicitly construct these systems, particularly for non-local objects and arbitrary spins.