QFT as a set of ODEs: higher dimensions

arXiv:2607.00079 · hep-th, hep-ph · Submitted 2026-06-30 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: High Energy Physics: Theory. The day in high-energy theory: gravity, strings, fields and the questions left open.

Kai: I'm Kai.

Mira: And I'm Mira, here with Lev.

Kai: We're looking at the paper: "QFT as a set of ODEs: higher dimensions".

Title and authors: Kai: The paper is titled "QFT as a set of ODEs: higher dimensions," and it brings together Fabiana De Cesare, Manuel Loparco, and the Department of Physics at the University of Turin.

Mira: I like that title because it suggests that what we usually think of as complex correlation functions can be reduced to a set of ordinary differential equations governing these fundamental numbers.

Lev: It sounds like they're taking the structure encoded in boundary operators and making it dynamic under changes in the bulk theory, which is what we need for real physics.

Kai: The core idea is that you start with the boundary data—the scaling dimensions and those OPE coefficients—and those numbers follow specific rules as you deform the bulk coupling.

Mira: And they show these rules are different for AdS three and AdS four, which is a significant step up from just two dimensions.

Lev: It moves the problem from being purely theoretical to actually having a concrete framework that can be used to model how these theories behave under deformation.

The paper's summary: Kai: What the paper does is derive these generalized flow equations by starting with correlation functions at finite coupling, using bulk perturbation theory and renormalization to link changes in those correlation functions to shifts in the bulk coupling.

Mira: They show that these flow equations are expressed as two specific differential equations involving the scaling dimensions, d and those bulk-boundary coefficients, which we call b phi i hat.

Lev: The first equation governs how fast the scaling dimensions change with respect to the coupling lambda, and it involves those integrals of conformal blocks we call I and J.

Kai: The second equation describes the flow of those bulk-boundary coefficients, showing how they evolve based on combinations of those same integrals and the OPE coefficients.

Mira: What’s really compelling is that these equations natively describe merger-annihilation when a scaling dimension hits that marginality value d, and level repulsion when different scaling dimensions approach each other in the flow.

The paper's improvements: Lev: They propose a few big methodological shifts to make this framework more practical for calculation, starting with substituting the ODE for those OPE coefficients with a crossing equation.

Kai: That makes sense because solving the crossing equation is way cheaper than solving the ordinary differential equations when you move up to higher dimensions, where those integrals get complicated very fast.

Mira: And they introduce Padé approximants to handle sums over boundary operators, which they say dramatically enhances convergence, especially for free theories.

Lev: That's a nice idea because standard local blocks require integrating these ODEs, which can be numerically expensive in higher dimensions than just solving a crossing equation.

Kai: They also check these things in free theories—like scalar and spinning operators—and those checks confirm that the Padé approximants work better than the standard or local blocks for those specific cases.

Conclusion: Mira: So, to wrap up, this paper shows that QFT data completely fixes the correlation functions in AdS space through these flow equations derived from bulk deformations.

Kai: The key insight is seeing merger-annihilation and level repulsion as natural features of these flow equations when you look at the scaling dimensions and their evolution.

Lev: From an engineering standpoint, it suggests that using crossing equations instead of solving the full ODEs for OPE coefficients is a much more efficient way to handle those calculations in higher dimensions.

Mira: And they point out that Padé approximants are a useful tool for ensuring better convergence when we are dealing with sums over boundary operators in free theories.

Kai: Ultimately, this work lays out the flow equations for QFT as a set of ODEs: higher dimensions, which gives us a dynamic way to understand how these theories react to changes in the bulk.

Lev: It opens up possibilities for studying things like glueball masses in three dimensional Yang-Mills and QCD in four dimensions, provided we can manage the required input data.

Mira: We'll keep an eye on those open questions they mention about finding a minimal set of QFT data and proving asymptotic sign modulation for generic interacting theories.

Istituto Nazionale di Fisica Nucleare Sezione di Torino · Department of Physics University of Turin

hep-th, hep-ph

Submitted: 2026-06-30

Updated: 2026-11-17

Comments: v1: 58 pages + 5 appendices, 18 figures, one ancillary Mathematica file; v2: added a discussion of the exclusion of the identity operator from the flow equations; revised the section on Padé approximation; corrected typos; v3: ancillary Mathematica file reinserted; v4: revisions after peer review

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Importance score: 80/100

The gist: Correlation functions of local operators in Quantum Field Theory (QFT) in Anti-de Sitter space (AdS) are completely determined by the QFT data…

Key concepts

Scaling Dimensions ($\Delta_i$)
These are parameters that characterize boundary operators in the QFT. They dictate how the correlation functions behave under changes in the bulk theory, and their proximity to specific values signals critical physical events like merger-annihilation.
Flow Equations
These are generalized ordinary differential equations that describe the evolution of QFT data (like scaling dimensions and boundary coefficients) when a bulk deformation is applied. They reveal how physical properties change dynamically as the theory is modified.
Merger-Annihilation and Level Repulsion
These are specific dynamical behaviors predicted by the flow equations. Merger-annihilation occurs when a boundary operator's scaling dimension hits a marginal value, while level repulsion describes how different scaling dimensions avoid getting too close to each other.
Padé Approximants
These are mathematical tools used to improve the accuracy of calculations involving sums over many boundary operators. They are particularly effective in free theories and ensure that the numerical solutions for the flow equations converge much faster than standard methods.

Terminology

Summary

Correlation functions of local operators in Quantum Field Theory (QFT) in Anti-de Sitter space (AdS) are completely determined by the QFT data, specifically the set of scaling dimensions delta i and Operator Product Expansion (OPE) coefficients C ijk of boundary operators, along with bulk-boundary (BOE) coefficients b phi i hat that encode how bulk fields decompose into boundary operators. p. 0 The work generalizes the ordinary differential equations (ODEs) governing the variation of this QFT data under a bulk relevant deformation, initially derived for AdS 2, to cases involving AdS 3 and AdS 4. p. 0 These generalized ODEs, termed flow equations, natively capture physical mechanisms such as merger-annihilation when a boundary operator approaches marginality (delta i = d) and level repulsion when different scaling dimensions delta i 's get close.

The paper derives these flow equations by starting from correlation functions at finite coupling lambda, utilizing bulk perturbation theory and renormalization to relate changes in correlation functions to shifts in the bulk coupling delta lambda. p. 0 The generalized ODEs in AdS d+1 are expressed as:

d delta i over d lambda = X l b phi l hat C(0) lii I(delta l),

db phi i hat over d lambda = X l X j b phi l hat b phi j hat C lijJ delta i (delta l, delta j),

where I and J are regulated integrals of conformal blocks.

A key methodological advancement is the proposal to substitute the ODE for OPE coefficients with a crossing equation to improve computational efficiency, as integrated blocks (K) required for the ODE are numerically expensive in higher dimensions, while solving crossing equations is relatively cheap. p. 5 Furthermore, Padé approximants are introduced to dramatically enhance the convergence of sums over boundary operators, particularly in free theories. p. 0

The flow equations exhibit specific dynamical properties: merger-annihilation occurs when delta i hits marginality (delta i = d), and level repulsion is observed as different delta i 's approach each other. The analysis shows that scaling dimensions near merger-annihilation follow a square root behavior, and BOE coefficients display similar characteristics. p. 0 For free theories, checks involving scalar and spinning operators confirm the efficacy of Padé approximants over standard or local blocks in this context. p. 7

The paper applies these flow equations to various QFTs in 3D and 4D, including Yang-Mills in 3D and QCD in 4D. p. 7 The minimal set of QFT data required includes locality sum rules that impose constraints among BOE and OPE coefficients. p. 4 Open questions remain regarding the minimal data set, proving asymptotic sign modulation for large scaling dimensions in generic interacting QFTs, and deriving closed-form ODEs for OPE coefficients. p. 0 Potential applications include studying glueball masses in 3D Yang-Mills and QCD in 3D, as well as investigating the conformal window of QCD. p. 3 The flow equations are valid when using local blocks or Padé approximants to ensure convergence. p. 7

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. The AI system can be upgraded to solve differential equations governing QFT data flow in Anti-de Sitter space by implementing a method that utilizes Pad´e approximants for standard conformal blocks, which are shown to converge fastest in free theories.

  2. The system will gain the ability to accurately predict merger-annihilation and level repulsion phenomena in strongly coupled QFTs by solving the flow equations (1.3) using crossing equations (1.4 or 1.5) instead of standard ODEs for OPE coefficients, which is proposed as a more efficient numerical approach.

  3. The AI can perform robust numerical analysis of QFT data by implementing algorithms to solve linear optimization problems, such as those used to determine OPE coefficients from crossing equations in AdS2 and AdS3, using methods like LinearOptimization on Mathematica.

  4. The system will be able to handle the convergence issues of flow equations by employing sequential Pad´e approximants for the double series involved in the flow of bulk-boundary-boundary coefficients, ensuring better accuracy than standard local block truncations.

  5. The AI can perform iterative procedures to determine the evolution of bulk operator boundary coefficients under a relevant deformation, using an iterative scheme like (7.5) to handle violations of assumptions regarding UV scaling dimensions and ensure finite flow equations without introducing an explicit UV cutoff.

Sources

Related papers