The Numerically Optimized Regulator and the Functional Renormalization Group
We aim to optimize the functional form of the compactly supported smooth (CSS) regulator within the functional renormalization group (RG), in the framework of bosonized two-dimensional quantum electrodynamics (QED2) and of the three-dimensional O(N = 1) scalar field theory in the local potential approximation (LPA). The principle of minimal sensitivity (PMS) is used for the optimization of the CSS regulator, recovering all the major types of regulators in appropriate limits. Within the investigated class of functional forms, a thorough investigation of the CSS regulator, optimized with two different normalizations within the PMS method, confirms that the functional form of a regulator first proposed by Litim is optimal within the LPA. However, Litim's exact form leads to a kink in the regulator function. A form of the CSS regulator, numerically close to Litim's limit while maintaining infinite differentiability, remains compatible with the gradient expansion to all orders. A smooth analytic behavior of the regulator is ensured by a small, but finite value of the exponential fall-off parameter in the CSS regulator. Consequently, a compactly supported regulator, in a parameter regime close to Litim's optimized form, but regularized with an exponential factor, appears to have favorable properties and could be used to address the scheme dependence of the functional RG, at least within the approximations employed in the studies reported here.
I. G. Marian et al., "The Numerically Optimized Regulator and the Functional Renormalization Group," Journal of Physics G: Nuclear and Particle Physics, vol. 41, no. 5, IOP Publishing, May 2014.
The definitive version is available at https://doi.org/10.1088/0954-3899/41/5/055001
Keywords and Phrases
CSS Regulator; Functional RG Flow; Numerical Calculation In The RG; Optimization Of Regulators In Nonperturbative Approaches To The Renormalization Group; Wetterich Equation
International Standard Serial Number (ISSN)
Article - Journal
© 2014 IOP Publishing, All rights reserved.