MEM study of true flattening of free energy and the $θ$ term

Masahiro Imachi,Yasuhiko Shinno,Hiroshi Yoneyama

DOI: https://doi.org/10.1016/j.nuclphysbps.2004.11.257

2004-10-05

Abstract:We study the sign problem in lattice field theory with a $\theta$ term, which reveals as flattening phenomenon of the free energy density $f(\theta)$. We report the result of the MEM analysis, where such mock data are used that `true' flattening of $f(\theta)$ occurs. This is regarded as a simple model for studying whether the MEM could correctly detect non trivial phase structure in $\theta$ space. We discuss how the MEM distinguishes fictitious and true flattening.

High Energy Physics - Lattice

What problem does this paper attempt to address?

This paper attempts to solve the sign problem encountered when introducing the θ - term in lattice field theory, especially the flattening phenomenon of the free - energy density \( f(\theta) \). Specifically, the author focuses on how to distinguish between false and real free - energy flattening and verify whether the maximum entropy method (MEM) can correctly detect the non - trivial phase structure.

### Main problems

1. **Sign problem and free - energy flattening**:
- In the lattice field theory containing the θ - term, a sign problem will occur during the calculation process.
- The traditional solution is to calculate the topological charge distribution \( P(Q) \) through Fourier transform, but this will lead to the flattening of the free - energy density \( f(\theta) \), thus misleading us to identify false first - order phase transition signals.

2. **Application of the maximum entropy method**:
- To overcome the above problems, the author uses the maximum entropy method (MEM) to analyze data.
- The research purpose is to verify whether MEM can correctly distinguish the false flattening caused by statistical errors from the real flattening caused by the characteristics of the data itself.

### Research methods

- **Model construction**:
- A simple model is constructed, in which the topological charge distribution \( P(Q) \) has a slight deviation at \( Q = 0 \), that is:
\[
P(Q)=P_G(Q)+p_0\times\delta_{Q, 0}
\]
- Where \( P_G(Q) \) is a Gaussian distribution and \( p_0 \) is a constant.

- **Data analysis**:
- Use numerical Fourier transform and MEM to analyze the generated simulation data.
- Compare the results under different \( p_0 \) values to determine whether MEM can correctly reproduce the real flattening.

### Results

- **Real flattening vs. false flattening**:
- For the case of \( p_0\neq0 \), MEM successfully reproduces the real flattening.
- For the case of \( p_0 = 0 \), the flattening is caused by statistical errors, and MEM cannot predict this false flattening.

- **Critical parameter \( p_0 \)**:
- When \( p_0\geq2.0\times 10^{-3} \), MEM can correctly reproduce the flattening.
- This critical value is related to the error magnitude in \( P(Q) \).

### Summary

This research verifies that MEM can effectively distinguish between false and real free - energy flattening by applying MEM to the simulation data with real flattening. This provides a powerful tool for understanding the non - trivial phase structure in lattice field theory.