Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2502.05504 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914273990541312 |
|---|---|
| author | Zhu, Qianteng Aarts, Gert Wang, Wei Zhou, Kai Wang, Lingxiao |
| author_facet | Zhu, Qianteng Aarts, Gert Wang, Wei Zhou, Kai Wang, Lingxiao |
| contents | We develop diffusion models for simulating lattice gauge theories, where stochastic quantization is explicitly incorporated as a physical condition for sampling. We demonstrate the applicability of this novel sampler to U(1) gauge theory in two spacetime dimensions and find that a model trained at a small inverse coupling constant can be extrapolated to larger inverse coupling regions without encountering the topological freezing problem. Additionally, the trained model can be employed to sample configurations on different lattice sizes without requiring further training. The exactness of the generated samples is ensured by incorporating Metropolis-adjusted Langevin dynamics into the generation process. Furthermore, we demonstrate that this approach enables more efficient sampling of topological quantities compared to traditional algorithms such as Hybrid Monte Carlo and Langevin simulations. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_05504 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Physics-Conditioned Diffusion Models for Lattice Gauge Theory Zhu, Qianteng Aarts, Gert Wang, Wei Zhou, Kai Wang, Lingxiao High Energy Physics - Lattice Machine Learning We develop diffusion models for simulating lattice gauge theories, where stochastic quantization is explicitly incorporated as a physical condition for sampling. We demonstrate the applicability of this novel sampler to U(1) gauge theory in two spacetime dimensions and find that a model trained at a small inverse coupling constant can be extrapolated to larger inverse coupling regions without encountering the topological freezing problem. Additionally, the trained model can be employed to sample configurations on different lattice sizes without requiring further training. The exactness of the generated samples is ensured by incorporating Metropolis-adjusted Langevin dynamics into the generation process. Furthermore, we demonstrate that this approach enables more efficient sampling of topological quantities compared to traditional algorithms such as Hybrid Monte Carlo and Langevin simulations. |
| title | Physics-Conditioned Diffusion Models for Lattice Gauge Theory |
| topic | High Energy Physics - Lattice Machine Learning |
| url | https://arxiv.org/abs/2502.05504 |