CERN Accelerating science

Preprint
Report number	CERN-TH-2025-204 ; arXiv:2510.18404
Title	Multi-Instantons, Multi-Axions, and Non-Invertible Symmetries in 4d QFT
Author(s)	Hong, Sungwoo (KAIST, Taejon) ; Kim, Hyungyu (KAIST, Taejon) ; Lee, Sung Mook (KAIST, Taejon ; CERN) ; Seo, Dongmin (KAIST, Taejon)
Document contact	Contact: arXiv
Imprint	2025-10-21
Number of pages	57
Note	57 pages, 3 figures
Subject category	hep-th ; Particle Physics - Theory
Abstract	We study non-invertible global symmetries in 4d quantum field theories, aiming to generalize existing discussions to theories with multiple instantons and axions, and to make the subject more accessible to particle phenomenology. Building on both the Adler-Bell-Jackiw (ABJ) anomaly construction and the half-space gauging approach, we identify the 3d topological quantum field theories required to describe non-invertible symmetries in the presence of multi-instanton effects. To this end, we introduce a method we call partial gauging and show that partial gauging of 3d Chern-Simons theories naturally leads to anomaly inflow actions with general ABJ anomaly matrices. We further generalize the half-space gauging construction to the case of multiple gauge sectors and compute correlation functions of boundary line operators. This enables us to analyze the action of non-invertible operators on various species of 't Hooft lines and to interpret the results in terms of the Witten effect. For theories with multiple axions, we determine both the non-invertible 0-form and 1-form symmetries, as well as their actions on axion strings and 't Hooft lines, thereby generalizing the previously known notion of the non-invertible Gauss law. Our framework unifies several previously disparate constructions, provides a concrete Lagrangian-based formulation of non-invertible symmetries, and naturally extends to theories with multiple instantons and axions, which are of both phenomenological and theoretical interest. These results may also find broader applications in beyond-the-Standard-Model scenarios.
Other source	Inspire
Copyright/License	preprint: (License: CC BY 4.0)