# Signal\-to\-noise improvement through neural network contour deformations for 3D 𝑺𝑼\(2\) lattice gauge theory

**INSPIRE:** [2692810](https://inspirehep.net/literature/2692810)
**arXiv:** [2309\.00600](https://arxiv.org/abs/2309.00600)
**DOI:** [10\.22323/1\.453\.0043](https://doi.org/10.22323/1.453.0043)

**Authors:** Lin, Yin, Detmold, William, Kanwar, Gurtej, Shanahan, Phiala E\., Wagman, Michael L\.

**Submitted:** 1 September 2023

**Subjects:**
- hep\-lat
- Lattice

**Journal reference:** PoS LATTICE2023 043 \(2024\)

**Citations:** 11

## Abstract

Complex contour deformations of the path integral have been demonstrated to significantly improve the signal\-to\-noise ratio of observables in previous studies of two\-dimensional gauge theories with open boundary conditions\. In this work, new developments based on gauge fixing and a neural network definition of the deformation are introduced, which enable an effective application to theories in higher dimensions and with generic boundary conditions\. Improvements of the signal\-to\-noise ratio by up to three orders of magnitude for Wilson loop measurements are shown in SU\(2\) lattice gauge theory in three spacetime dimensions\.