Fixed-point tensor is a four-point function

Atsushi Ueda,Masahito Yamazaki

2023-08-04

Abstract:Through coarse-graining, tensor network representations of a two-dimensional critical lattice model flow to a universal four-leg tensor, corresponding to a conformal field theory (CFT) fixed-point. We computed explicit elements of the critical fixed-point tensor, which we identify as the CFT four-point function. This allows us to directly extract the operator product expansion coefficients of the CFT from these tensor elements. Combined with the scaling dimensions obtained from the transfer matrix, we determine the complete set of the CFT data from the fixed-point tensor for any critical unitary lattice model.

High Energy Physics - Theory,Statistical Mechanics,Strongly Correlated Electrons

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to accurately extract the fixed - point tensor from the tensor network representation, and then determine all the data of the corresponding conformal field theory (CFT), including the central charge, scaling dimension and operator product expansion (OPE) coefficient. Specifically, through the tensor network renormalization (TNR) method, the author studied the behavior of two - dimensional critical lattice models during the coarse - graining process and found that the tensor network representations of these models would converge to a general four - legged tensor, which corresponds to the fixed point of the CFT.

### Main Contributions
1. **Determination of Four - Legged Fixed - Point Tensor**:
- The author calculated the explicit elements of the critical fixed - point tensor and related them to the four - point function of the CFT. This enables the extraction of the CFT's OPE coefficients directly from these tensor elements.
- By combining the scaling dimension obtained from the transfer matrix, the author was able to determine the complete CFT data for any critical unitary lattice model.

2. **Analytical Expressions**:
- The author derived the analytical expressions of the three - legged tensor \( S^*_{\alpha\beta\gamma} \) and the four - legged tensor \( T^*_{\alpha\beta\gamma\delta} \) and proved that they are proportional to the three - point function and the four - point function of the CFT respectively:
\[
\frac{S^*_{\alpha\beta\gamma}}{S^*_{111}}=\langle\phi_\alpha(-x_S)\phi_\beta(ix_S)\phi_\gamma(0)\rangle_{pl}
\]
\[
\frac{T^*_{\alpha\beta\gamma\delta}}{T^*_{1111}}=\langle\phi_\alpha(-x_T)\phi_\beta(ix_T)\phi_\gamma(x_T)\phi_\delta(-ix_T)\rangle_{pl}
\]
where \( x_S = e^{\pi/4} \) and \( x_T = e^{\pi/2}/2 \).

3. **Numerical Verification**:
- The author carried out numerical verification using the tensor renormalization group (TRG) method and confirmed the consistency between the theoretical values and the numerical results. In particular, for the Ising model and the 3 - state Potts model, the results of the numerical calculations are in line with the theoretical expectations.

### Motivation
1. **Extraction of CFT Data**:
- The fixed - point tensor encodes the data of the CFT, including the central charge and the scaling dimension. Calculating the OPE coefficient is a complex problem, and the author's method provides a new way to solve this problem.

2. **Realization of RG Flow Specification**:
- Determining the fixed - point tensor helps to specifically realize the RG flow, especially the behavior near the critical point. This helps to understand the stability of the fixed point.

3. **Evaluation of Algorithm Precision**:
- The exact fixed - point tensor expression can be used as a benchmark for evaluating the precision of different tensor network algorithms.

### Conclusion
By introducing an accurate tensor network representation, the author has solved the problem of numerically determining the critical RG fixed point, provided an important tool for the practical calculation of fixed - point theories, and laid the foundation for the strict verification of RG theory.