Applications of CFT determinant formulas in number theory

Basic Information

• Paper ID: 2508.01301
• Title: Applications of CFT determinant formulas in number theory
• Authors: D. Levin (University of Oxford), H.-G. Shin (Charles University), A. Zuevsky (Czech Academy of Sciences)
• Classification: math.FA (Functional Analysis), with cross-disciplinary connections to 17B69 (Vertex operator algebras), 30F10 (Compact Riemann surfaces), 32A25 (Integral representations), 11F03 (Modular and automorphic functions)
• Publication Date: arXiv v3 version, November 17, 2025
• Paper Link: https://arxiv.org/abs/2508.01301v3

Abstract

This paper demonstrates how to utilize determinant representations of correlation functions in conformal field theory (CFT) to derive new determinant formulas for powers of the modular discriminant, expressed through parametrized deformed elliptic functions. In particular, the authors obtain a generalization of the Garvan formula corresponding to the genus 2 Riemann surface case.

Research Background and Motivation

Research Problem

The core problem addressed in this paper is: How can determinant representation techniques from conformal field theory (CFT) be applied to number theory, particularly to derive explicit determinant formulas for higher powers of the modular discriminant $\Delta(τ) = η(τ)^{24}$?

Problem Significance

1. Interdisciplinary Connection: Many identities in number theory originate from correlation function calculations in conformal field theory, revealing deep relationships between modular forms, fundamental kernels, and q-series.
2. Generalization of Classical Results: The classical Garvan formula provides a 3×3 determinant representation for $\Delta^2(τ)$, but formulas for higher powers and higher genus cases remain lacking.
3. Theoretical Value: These formulas connect vertex operator algebras, Riemann surface theory, and modular form theory, providing new tools for understanding the deep structures of these fields.

Limitations of Existing Methods

1. Restrictions of the Garvan Formula: The classical Garvan formula only handles the case of $\Delta^2(τ)$, lacking systematic methods for arbitrary powers $\Delta^n(τ)$.
2. Genus Limitations: Existing formulas primarily address genus 1 (torus) cases, with corresponding results for higher genus Riemann surfaces still undeveloped.
3. Missing Parametrization: Classical formulas use standard Eisenstein series, lacking the flexibility of parametric deformation.

Research Motivation

The authors systematically derive determinant formulas for arbitrary powers of the modular discriminant using correlation function calculations from vertex operator algebras on tori and genus 2 Riemann surfaces. Through bosonization-fermionization duality, they introduce deformed Weierstrass functions and Eisenstein series as matrix elements.

Core Contributions

1. Generalization of Garvan Formula to Arbitrary Powers: For $\Delta^n(τ)$ (arbitrary positive integer $n$), two classes of determinant representations are provided (Proposition 1):
   • For $(θ,φ) ≠ (1,1)$: $8n×8n$ matrix determinant
   • For $(θ,φ) = (1,1)$: $(8n+1)×(8n+1)$ matrix determinant
2. Introduction of Deformed Elliptic Function Framework: Systematic use of parametrized deformed Weierstrass functions $P_1^{(1)}[^θ_φ](z,τ)$ and deformed Eisenstein series $E_n^{(1)}[^θ_φ](τ)$, providing more flexible representations.
3. Establishment of Genus 2 Riemann Surface Formulas: Through self-sewing techniques, determinant representations for $η^{3κ²}(τ)$ in the genus 2 case are derived (Proposition 3), involving genus 2 theta functions and Szegő kernels.
4. General Formulas for Higher Powers: For $\Delta^{8Φ}(τ)$ (Proposition 2), determinant representations involving block matrices are provided, with matrix elements constructed from second-order derivative operators $D^{(1)}[^θ_φ](k,l,z,τ)$.
5. Systematic Connection Between CFT and Number Theory: Explicitly demonstrates how to derive number-theoretic identities from correlation functions of vertex operator algebras and elliptic versions of Fay's trisecant identity.

Methodology Details

Theoretical Framework

The core method is based on correlation function calculations from vertex operator algebras (VOA) on Riemann surfaces, particularly utilizing boson-fermion duality.

Torus Correlation Functions: For the twisted module by automorphism $g$ of vertex operator algebra $V$, the correlation function on the torus is defined as: $Z_V^{(1)}\left[\begin{matrix}f\\g\end{matrix}\right](v_1,z_1,\ldots,v_n,z_n;q) = \text{STr}_V\left(fY(v_1,z_1)\cdots Y(v_n,z_n)q^{L(0)-C/24}\right)$

where $L(0)$ is the Virasoro algebra generator, $C$ is the central charge, and $q=e^{2πiτ}$.

Deformed Elliptic Functions

Deformed Weierstrass Function is defined as: $P_1^{(1)}\left[\begin{matrix}θ\\φ\end{matrix}\right](z,τ) = -\sum'_{n∈\mathbb{Z}+λ}\frac{q^nz}{1-θ^{-1}q^n} = \frac{1}{z} - \sum_{n≥1}\frac{1}{n}E_n^{(1)}\left[\begin{matrix}θ\\φ\end{matrix}\right](τ)z^{n-1}$

where the prime indicates omission of the $n=0$ term when $(θ,φ)=(1,1)$.

Deformed Eisenstein Series: $E_n^{(1)}\left[\begin{matrix}θ\\φ\end{matrix}\right](τ) = -\frac{B_n(λ)}{n!} + \frac{1}{(n-1)!}\sum'_{r≥0}\frac{(r+λ)^{n-1}θ^{-1}q^{r+λ}}{1-θ^{-1}q^{r+λ}} + \frac{(-1)^n}{(n-1)!}\sum_{r≥1}\frac{(r-λ)^{n-1}θq^{r-λ}}{1-θq^{r-λ}}$

where $θ=e^{-2πiα}$, $φ=e^{-2πiβ}$, and $λ$ is related to $α,β$.

Core Technique: Elliptic Version of Fay's Trisecant Identity

For the matrix $P_n(θ,φ) = [P_1^{(1)}[^θ_φ](x_i-y_j,τ)]_{1≤i,j≤n}$, the paper proves:

Case 1 $(θ,φ)≠(1,1)$: $\det P_n(θ,φ) = \frac{ϑ^{(1)}[^{1/2+α}_{1/2-β}](\sum_{i=1}^n(x_i-y_i),τ)}{ϑ^{(1)}[^{1/2+α}_{1/2-β}](0,τ)}K_n^{(1)}(\mathbf{x},\mathbf{y})$

Case 2 $(θ,φ)=(1,1)$: For the augmented matrix $Q_n(τ)$ (adding a row and column of 1's and 0's to $P_n(1,1)$): $\det Q_n = -K^{(1)}\left(\sum_{i=1}^n(x_i-y_i),τ\right)K_n^{(1)}(\mathbf{x},\mathbf{y})$

where $K^{(1)}(z,τ) = -\frac{i}{η^3(τ)}ϑ^{(1)}[^{1/2}_{1/2}](z,τ)$ is the genus 1 prime form, and $K_n^{(1)}(\mathbf{x},\mathbf{y}) = \frac{\prod_{1≤i<j≤n}K^{(1)}(x_i-x_j,τ)K^{(1)}(y_i-y_j,τ)}{\prod_{1≤i,j≤n}K^{(1)}(x_i-y_j,τ)}$

Derivation Strategy

1. Starting from Free Fermion VOA: Calculate two representations of the twisted partition function (fermion and boson pictures)
2. Comparing Coefficients: By comparing theta functions and infinite products in both representations, obtain generalizations of the Jacobi triple product formula
3. Applying Fay Identity: Use the elliptic Fay identity to connect determinants with prime forms and theta functions
4. Parameter Specialization: Choose specific parameter values (e.g., $α=β=1/2$) to obtain modular discriminant formulas

Self-Sewing Technique for Genus 2

For genus 2 Riemann surfaces, obtained by sewing two tori along boundaries. Key techniques include:

1. Genus 2 Partition Function Decomposition: $Z^{(2)}\left[\begin{matrix}f\\g\end{matrix}\right](τ_1,τ_2,ϵ) = Z^{(1)}\left[\begin{matrix}f_1\\g_1\end{matrix}\right](τ_1)Z^{(1)}\left[\begin{matrix}f_2\\g_2\end{matrix}\right](τ_2)\det(I-Q^{(1)})^{1/2}$
2. Genus 2 Szegő Kernel: $S^{(2)}(x,y) = S_κ^{(2)}(x,y) + ξh(x)D(θ)(I-T^{(2)})^{-1}h^t(y)$

where the sewing parameter $κ$ and complex matrix structures are involved.

3. Final Formula (Proposition 3): $η^{3κ^2}(τ) = \frac{e^{-2iπβ_2κ}ϑ^{(2)}[^α_β](Ω^{(2)})}{ϑ^{(1)}[^{1/2}_{1/2}](w,τ)^{κ^2}}\frac{\det[S_n^{(2)}[^α_β](\mathbf{x},\mathbf{y})](-e^{iπB}ρ)^{\frac{1}{2}κ^2}}{ϑ^{(1)}[^{α_a}_{β_a}](κw,τ)\det(I-R)^{1/2}}\det\begin{bmatrix}S_{κ,n}^{(2)} & -ξHD(θ_2)\\H^t & I-T\end{bmatrix}$

Experimental Setup

Note: This is a pure mathematics theory paper with no experimental verification or numerical computation. The authors explicitly state at the beginning:

• No datasets used
• No data analysis performed
• No AI involvement in writing

Therefore, "verification" in this paper is primarily through:

1. Rigor of Mathematical Proof: Based on established vertex operator algebra and Riemann surface theory
2. Consistency with Known Results: Special cases reduce to the classical Garvan formula
3. Self-Consistency of Theoretical Framework: Logical connections between propositions

Main Results

Proposition 1: Modular Discriminant Formulas for Arbitrary Powers

Result 1a (Case $(θ,φ)≠(1,1)$): $\Delta^n(τ) = -\frac{ϑ^{(1)}[^{1/2}_{1/2}](0,τ)Θ_{8n,8n,(1,1)}^{(1)}(\mathbf{x},\mathbf{y},τ)}{ϑ^{(1)}[^{1/2-β}_{1/2+α}](\sum_{i=1}^{8n}(x_i-y_i),τ)}\det P_{8n}(θ,φ)$

where $Θ_{r,s,(m_i,n_j)}^{(1)}$ is the product of theta functions: $Θ_{r,s,(m_i,n_j)}^{(1)} = \frac{\prod_{i,j}ϑ^{(1)}[^{1/2}_{1/2}](x_i-y_j,τ)^{m_in_j}}{\prod_{i<k}ϑ^{(1)}[^{1/2}_{1/2}](x_i-x_k,τ)^{m_im_k}\prod_{j<l}ϑ^{(1)}[^{1/2}_{1/2}](y_j-y_l,τ)^{n_jn_l}}$

Result 1b (Case $(θ,φ)=(1,1)$): $\Delta^n(τ) = i\frac{Θ_{8n+1,8n+1,(1,1)}^{(1)}(\mathbf{x},\mathbf{y},τ)}{ϑ^{(1)}[^{1/2}_{1/2}](\sum_{i=1}^{8n+1}(x_i-y_i),τ)}\det Q_{8n+1}$

Significance:

• Represents $\Delta^n(τ)$ as an $8n×8n$ (or $(8n+1)×(8n+1)$) determinant
• Matrix elements are deformed Weierstrass functions
• Normalized through theta function ratios

Proposition 2: Block Matrix Formula for High Powers

For $ζ=8Φ$, where $Φ = \sum_{i,k}m_in_j - \sum_{i<k}m_im_k - \sum_{j<l}n_jn_l$:

$\Delta^ζ(τ) = (-i)^{Φ/24}\frac{ϑ^{(1)}[^{1/2-β}_{1/2+α}](0,τ)Θ_{r,s,(\mathbf{m},\mathbf{n})}^{(1)}(\mathbf{x},\mathbf{y},τ)}{ϑ^{(1)}[^{1/2-β}_{1/2+α}](\sum_im_ix_i-\sum_jn_jy_j,τ)}\det D_{r,s}$

where $D_{r,s}$ is a block matrix with blocks $D^{(ab)}$ being $m_a×n_b$ matrices: $D^{(ab)}(i,j) = D^{(1)}\left[\begin{matrix}θ\\φ\end{matrix}\right](i,j,x_a-y_b,τ)$

and $D^{(1)}\left[\begin{matrix}θ\\φ\end{matrix}\right](k,l,z,τ) = (-1)^{k+1}\binom{k+l-2}{k-1}P_{k+l-1}^{(1)}\left[\begin{matrix}θ\\φ\end{matrix}\right](τ,z)$

Significance:

• Provides a more general framework handling more complex power structures
• Block matrix structure reflects combinatorial properties of multi-point correlation functions

Proposition 3: Genus 2 Formula

$η^{3κ^2}(τ) = \frac{e^{-2iπβ_2κ}ϑ^{(2)}[^α_β](Ω^{(2)})}{ϑ^{(1)}[^{1/2}_{1/2}](w,τ)^{κ^2}}\times\frac{\det[S_n^{(2)}[^α_β]](-e^{iπB}ρ)^{\frac{1}{2}κ^2}}{ϑ^{(1)}[^{α_a}_{β_a}](κw,τ)\det(I-R)^{1/2}}\det\begin{bmatrix}S_{κ,n}^{(2)} & -ξHD(θ_2)\\H^t & I-T\end{bmatrix}$

Significance:

• First Garvan-type formula for genus 2 case
• Involves genus 2 theta functions $ϑ^{(2)}$ and period matrix $Ω^{(2)}$
• Determinant couples finite and infinite-dimensional parts

Connection with Classical Results

Classical Garvan Formula ($n=1$): $\Delta(τ) = \frac{1}{1728}\det\begin{bmatrix}E_4^{(1)}(τ) & E_6^{(1)}(τ)\\E_6^{(1)}(τ) & E_8^{(1)}(τ)\end{bmatrix}$

$\Delta^2(τ) = -\frac{691}{250(1728)^2}\det\begin{bmatrix}E_4^{(1)} & E_6^{(1)} & E_8^{(1)}\\E_6^{(1)} & E_8^{(1)} & E_{10}^{(1)}\\E_8^{(1)} & E_{10}^{(1)} & E_{12}^{(1)}\end{bmatrix}$

The paper's formulas should reduce to the above classical formulas when $n=1,2$ and parameters take special values, though explicit verification is not provided.

Technical Innovations

1. Introduction of Deformation Parameters

• Innovation: Use of parametrized deformed function families $(θ,φ)$ rather than fixed classical Eisenstein series
• Advantages:
  • Provides greater flexibility through parameter adjustment
  • Reflects natural structure of twisted modules in VOA
  • Allows handling of degenerate case $(θ,φ)=(1,1)$

2. Application of Elliptic Fay Identity

• Innovation: Generalization of Fay's trisecant identity from algebraic geometry to elliptic function setting
• Technical Points:
  • Establishes relationship between determinants and prime forms
  • Encodes combinatorial information of points through $K_n^{(1)}(\mathbf{x},\mathbf{y})$
  • Utilizes addition formulas for theta functions

3. Self-Sewing Technique

• Innovation: Construction of higher genus formulas through torus self-sewing
• Technical Challenges:
  • Handling complex dependence on sewing parameters $ρ,κ$
  • Managing coupling of finite and infinite-dimensional matrices
  • Controlling branch cut parameters $B$

4. Unified Framework

• Innovation: Incorporation of formulas for different genera and powers into unified CFT framework
• Theoretical Value:
  • Reveals deep connections between modular form theory and VOA theory
  • Provides clear blueprint for further generalizations

Related Work

Historical Background

1. Jacobi Triple Product Formula: Earliest example connecting theta functions and infinite products
2. Kac (1974): Natural emergence of $η(τ)$ identities from affine Lie algebra studies
3. Garvan (unpublished): 3×3 determinant formula for $\Delta^2(τ)$
4. Milne (2001): Proof and generalization of Garvan formula

CFT and Number Theory Connections

• Dong, Li, Mason (2000): Introduction of deformed Weierstrass functions and Eisenstein series
• Mason, Tuite, Zuevsky (2008): Systematic development of correlation function theory for R-graded VOA on tori, establishing connections with deformed elliptic functions
• Zhu (1996): Modular invariance theory providing foundation for CFT methods

Riemann Surface Theory

• Fay (1973): Theta function theory and trisecant identities
• Mumford (1983): Tata Lectures, systematic theta function theory
• Tuite, Zuevsky (2011-2013): Development of Szegő kernel theory on sewn Riemann surfaces, establishing genus 2 fermion partition functions

Positioning of This Paper

• Relation to Garvan's Work: Direct generalization to arbitrary powers and higher genus
• Relation to CFT Literature: Application of techniques developed in 8,13, but focused on number-theoretic applications
• Relation to Modular Form Theory: Provides new determinant representations complementing classical theory

Conclusions and Discussion

Main Conclusions

1. Arbitrary Power Formulas: Successfully derives determinant representations for $\Delta^n(τ)$ for all $n≥1$, with matrix dimensions $8n×8n$ or $(8n+1)×(8n+1)$
2. Effectiveness of Deformed Functions: Parametrized deformed Eisenstein series and Weierstrass functions provide natural language for expressing modular discriminants
3. Genus 2 Generalization: First establishment of Garvan-type formulas on genus 2 Riemann surfaces, involving complex self-sewing structures
4. Power of CFT Methods: Demonstrates systematic effectiveness of vertex operator algebra techniques in deriving number-theoretic identities

Limitations

1. Missing Explicit Verification:
   • Paper does not provide explicit comparison with classical Garvan formula for $n=1,2$
   • Lacks concrete numerical examples with specific parameter choices
   • Optimal parameter selection criteria for $(θ,φ)$ remain unclear
2. Formula Complexity:
   • Involves extensive theta function products and ratios
   • Matrix dimensions grow rapidly with $n$ ($8n×8n$)
   • Genus 2 formula contains infinite-dimensional matrices, making practical computation difficult
3. Theoretical Completeness:
   • Uniqueness of formulas not discussed
   • Relationships between different parametrizations unclear
   • Generalization path for genus $g≥3$ not explicitly outlined
4. Application Scope:
   • Primarily focuses on modular discriminant; generalizations to other modular forms not explored
   • Relationship with Siegel modular forms (higher-dimensional generalizations) not established

Future Directions

The paper implicitly suggests the following research directions:

1. Higher Genus: Generalization to arbitrary genus $g$ using self-sewing techniques
2. Other Modular Forms: Application of methods to Eisenstein series themselves, Ramanujan $τ$-function, etc.
3. Numerical Algorithms: Development of efficient algorithms for computing these determinants
4. q-Series Identities: Derivation of new q-series identities from determinant formulas
5. Physical Applications: Applications in string theory and conformal field theory

In-Depth Evaluation

Strengths

1. Theoretical Innovation

• Original Contribution: First systematic derivation of Garvan-type formulas for arbitrary powers and genus 2, filling important theoretical gaps
• Novel Methods: Creative application of CFT techniques (particularly elliptic Fay identity) to number-theoretic problems
• Unified Framework: Establishes systematic path from VOA correlation functions to modular form identities

2. Mathematical Rigor

• Complete Proofs: Propositions 1-3 provide clear proof strategies
• Clear Logic: Natural progression from torus case to genus 2 generalization
• Adequate References: Appropriately cites previous technical work 8,13

3. Theoretical Depth

• Cross-Disciplinary Connections: Deeply reveals intrinsic relationships between algebraic geometry (Riemann surfaces), Lie theory (VOA), and number theory (modular forms)
• Generalization Potential: Provided framework has clear paths for continued generalization

4. Writing Quality

• Reasonable Structure: Clear hierarchy from motivation to results
• Clear Notation: Despite involving complex objects, notation system is largely consistent
• Sufficient Background: Appropriately reviews necessary background knowledge

Weaknesses

1. Practical Applicability Issues

• Computational Complexity: $8n×8n$ matrices are impractical to compute for large $n$
• Missing Algorithms: No effective algorithms provided for computing matrix elements
• Parameter Dependence: Formulas depend on auxiliary variables $\mathbf{x},\mathbf{y}$ with unclear selection criteria

2. Insufficient Verification

• No Numerical Examples: Should at least verify $n=1,2$ cases
• Missing Special Cases: Does not explicitly show how to reduce to classical formulas
• Incomplete Consistency Checks: Compatibility between different propositions insufficiently discussed

3. Technical Details

• Theta Function Properties: Specific properties of $Θ_{r,s,(m_i,n_j)}^{(1)}$ insufficiently elaborated
• Convergence: Convergence of infinite products and series not discussed
• Branch Cuts: Role of parameter $B$ in genus 2 formula not sufficiently clear

4. Presentation Issues

• Symbol Overload: Extensive use of superscripts and subscripts sometimes difficult to track
• Omitted Details: Some derivation steps overly abbreviated (e.g., "Then Proposition follows")
• Insufficient Motivation: Choice of specific matrix forms not adequately explained

Impact Assessment

Contribution to Academia

1. Number Theory: Provides new determinant representation tools for modular form theory
2. CFT: Demonstrates concrete applications of VOA techniques to number theory
3. Algebraic Geometry: Enriches theory of special functions on Riemann surfaces

Practical Value

• Theoretical Value: High for understanding modular form structure
• Computational Value: Limited due to complexity; direct computational applications restricted
• Inspirational Value: Strong; methods may inspire derivation of other modular form identities

Reproducibility

• Theoretical Reproducibility: Proofs based on published results, verifiable
• Computational Reproducibility: Poor; lacks specific algorithms and code
• Background Requirements: Understanding requires deep knowledge of VOA, Riemann surfaces, and modular forms

Applicable Scenarios

1. Theoretical Research

• Modular Form Theory: Study of modular discriminant and Eisenstein series structure
• Vertex Operator Algebras: Understanding VOA connections with number theory
• Algebraic Geometry: Special functions on Riemann surfaces

2. Potential Applications

• String Theory: Partition function calculations
• Conformal Field Theory: Explicit correlation function formulas
• Number Theory Algorithms: May inspire new computational methods (requires further development)

3. Educational Value

• Graduate Courses: Exemplar of interdisciplinary mathematics
• Seminar Topics: Case study of CFT methods in number theory

Key References

1. 8 Mason, Tuite, Zuevsky (2008): Core technical foundation of this paper, develops torus correlation function theory for R-graded VOA and deformed elliptic functions
2. 13 Tuite, Zuevsky (arXiv:1308.2441): Detailed genus 2 fermion partition function calculations, direct source for Proposition 3
3. 2 Fay (1973): Theta Functions on Riemann Surfaces, provides original Fay trisecant identity theory
4. 3 Garvan (unpublished): Source of original Garvan formula, starting point for this paper's generalizations
5. 6 Milne (2001): Proof and generalization of Garvan formula, provides theoretical background
6. 11 Tuite, Zuevsky (2011): Szegő kernel on sewn Riemann surfaces, foundation of self-sewing techniques

Overall Assessment

This is a theoretically profound but practically limited pure mathematics paper. The authors successfully apply powerful conformal field theory techniques to classical number-theoretic problems, deriving new determinant formulas for arbitrary powers of the modular discriminant and pioneering corresponding results for genus 2. The paper's main value lies in:

✅ Theoretical Breakthrough: Systematic generalization of Garvan formula, establishing unified framework
✅ Methodological Innovation: Demonstrates systematic application of CFT techniques to number theory
✅ Interdisciplinary Connections: Deepens relationships between VOA, Riemann surfaces, and modular forms

Yet notable deficiencies exist:

⚠️ Missing Verification: Lacks explicit comparison with classical results and numerical examples
⚠️ Poor Practicality: Formulas too complex for practical computation
⚠️ Insufficient Details: Some technical details and motivations inadequately explained

Recommended Readers: Professional researchers in vertex operator algebras, modular form theory, or algebraic geometry. Requires substantial mathematical background for full comprehension.

Academic Value: ⭐⭐⭐⭐☆ (4/5) - Important theoretical contribution, opens new directions
Practical Value: ⭐⭐☆☆☆ (2/5) - Currently primarily theoretical interest
Readability: ⭐⭐⭐☆☆ (3/5) - Requires high professional background
Impact Potential: ⭐⭐⭐⭐☆ (4/5) - May inspire important subsequent work