Comparison of |Q|=1 and |Q|=2 gauge-field configurations on the lattice four-torus
Sundance O. Bilson-Thompson, Derek B. Leinweber, Anthony G. Williams,, Gerald V. Dunne

TL;DR
This study investigates the instability of |Q|=1 gauge-field configurations on the lattice 4-torus for SU(3), identifying early signals of their decay and contrasting them with stable |Q|=2 configurations, using advanced error control techniques.
Contribution
It extends previous SU(2) work to SU(3), identifying global instability signals of |Q|=1 configurations on the lattice 4-torus with improved discretization error control.
Findings
|Q|=1 configurations show early instability signals before instanton shrinkage.
These signals are absent in |Q|=2 configurations, indicating a global instability.
Discretization errors are minimized, and topological charge fluctuations are observed.
Abstract
It is known that exactly self-dual gauge-field configurations with topological charge |Q|=1 cannot exist on the untwisted continuum 4-torus. We explore the manifestation of this remarkable fact on the lattice 4-torus for SU(3) using advanced techniques for controlling lattice discretization errors, extending earlier work of De Forcrand et. al. for SU(2). We identify three distinct signals for the instability of |Q|=1 configurations, and show that these manifest themselves early in the cooling process, long before the would-be instanton has shrunk to a size comparable to the lattice discretization threshold. These signals do not appear for our |Q|=2 configurations. This indicates that these signals reflect the truly global nature of the instability, rather than local discretization effects. Monte-Carlo generated SU(3) gauge field configurations are cooled to the self-dual limit using an…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
