A numerical study reports signs that a compact U(1) lattice model can be followed along a path where several low-lying mass ratios remain broadly stable while its lattice-unit string-tension scale falls. The pattern is consistent with a possible continuum limit, but it is not proof that such a limit exists.
The document is a preprint, not a settled result. Its new simulations use β values from 1.9 to 2.2, alongside a Wilson-action reference at β=1.8 and µ=0. Here β is the dimensionless lattice coupling and µ is an added coupling tied to the total number of lattice monopoles.
A second control knob
The central idea is to tune β and µ together instead of changing β alone. The researchers first anchored the path to the Wilson-action reference at β=1.8 and µ=0, where the lightest glueball-to-string-tension ratio was 2.317(15). They then interpolated the µ values needed at higher β to keep that normalized ratio at 2.317(15).
The fitted relation between the two couplings was . In plain language, the fitted monopole coupling rises with β. Its slope was described as close to an approximately 2.49 prediction from a dilute-monopole-gas picture, although the simulated regime itself was not dilute.
The numerical work used Monte Carlo sampling. Each update cycle combined one Metropolis sweep with ten pseudo-microcanonical sweeps, and measurements were taken every 50 cycles. The study collected approximately 5 × 10^4 measurements for each β–µ pair.
Ratios held broadly steady
The main test was whether the chosen path kept the lightest 0−− glueball-to-string-tension ratio fixed. Two other low-lying checks showed only limited movement across the same range: the first-excited 0−− ratio ran from 5.40(20) to 5.461(43), while the ground-state 0++ ratio ran from 3.952(27) to 4.02(10).
The authors judged both additional ratios consistent with being independent of β. That is an encouraging sign for a line of constant physics, meaning a path through the model’s parameter space where the measured relationships stay stable. The uncertainties were relatively large, however, so the result is better read as a broad consistency than as a sharply tested prediction.
The string-tension signal moved in the expected direction for a possible continuum approach. Its square root in lattice units fell from 0.2802(80) at β=1.9 to 0.2355(12) at β=2.2. The decline was slower than in the µ=0 Wilson action and looked approximately linear over the range studied.
A promising trend with a short runway
The authors caution that the apparent linear trend may be a finite-range effect. The study covers only β=1.9 to 2.2, and its data do not definitively establish either a critical point or a continuum limit. It also leaves open whether the required critical point exists at finite β or only as β approaches infinity.
There is also a tension in the comparison with the dilute-monopole estimate. On the symmetric lattices, the reported monopole density rose from approximately 0.78 at β=1.9 to approximately 0.96 at β=2.2, and the authors state that monopoles were not dilute in this regime. That makes the fitted slope’s closeness to the approximately 2.49 dilute-gas value suggestive, rather than a quantitatively reliable test of that approximation.
The checks were also limited in scope. The constant-physics path was defined using one lightest-glueball ratio and tested with two additional low-lying ratios. The string-tension analysis used ground-state winding flux tubes, while excited flux-tube states were not part of the analysis. Dedicated finite-volume tests were reported to show that finite-volume effects were well controlled for the relevant analysis.
The next test is farther out
For now, the data do not support a definitive continuum conclusion. The preprint presents a numerical proposal whose trends will need to be tested beyond the explored range. Appendix B contains the raw lattice data corresponding to Tables 2–5.
The document was prepared for submission to the Journal of High Energy Physics. CB and ISC acknowledge support from the 2022 PRIN project “Emerging gauge theories: critical properties and quantum dynamics”; the simulations used the CSN4 cluster at INFN-PISA and the Cyclone high-performance computing cluster at The Cyprus Institute.
Paper data and sources
Original title: A different kind of continuum limit for the three-dimensional U(1) gauge theory
Authors: Andreas Athenodorou, Claudio Bonati, Ivan Soler Calero
Journal/Repository: arXiv
Status: Preprint, not yet peer-reviewed
First online: 2026-08-28
DOI: Not available
Original paper · Full text