Searcharxiv⌕ Search

arXiv subjects

Urs Wenger

Publications and source records attributed to Urs Wenger.

At least 19 recordsLinked to original sources
arXiv2026

Higher Mellin Moments of the Unpolarized PDF of the Pion and the Kaon from Lattice QCD

We present results on the Mellin moments of the unpolarized parton distribution function (PDF) of the pion and kaon up to the fourth order. The computation is done using one $N_f=2+1+1$ gauge ensemble of twisted mass fermions with quark masses tuned to approximately their physical values. We reconstruct the valence pion and kaon PDFs using the connected contributions to the three Mellin moments. We compare our results on the Mellin moments and the reconstructed PDFs with other lattice QCD and phenomenological determinations.

hep-lat↗
arXiv2026

Machine learning for four-dimensional SU(3) lattice gauge theories

In this review I summarize how machine learning can be used in lattice gauge theory simulations and what ap\-proaches are currently available to improve the sampling of gauge field configurations, with a focus on applications in four-dimensional SU(3) gauge theories. These include approaches based on generative machine-learning models such as (stochastic) normalizing flows and diffusion processes, and an approach based on renormalization group (RG) transformations, more specifically the machine learning of RG-improved gauge actions using gauge-equivariant convolutional neural networks. In particular, I present scaling results for a machine-learned fixed-point action in four-dimensional SU(3) gauge theory towards the continuum limit. The results include observables based on the classically perfect gradient-flow scales, which are free of tree-level lattice artefacts to all orders, and quantities related to the static potential and the deconfinement transition.

hep-lat↗
arXiv2025

Machine-learned RG-improved gauge actions and classically perfect gradient flows

Extracting continuum properties of quantum field theories from discretized spacetime is challenging due to lattice artifacts. Renormalization-group (RG)-improved lattice actions can preserve continuum properties, but are in general difficult to parameterize. Machine learning (ML) with gauge-equivariant convolutional neural networks provides a way to efficiently describe such actions. We test a machine-learned RG-improved lattice gauge action, the classically perfect fixed-point (FP) action, for four-dimensional SU(3) gauge theory through Monte Carlo simulations. We establish that the gradient flow of the FP action is free of tree-level discretization effects to all orders in the lattice spacing, making it classically perfect. This allows us to test the quality of improvement of the FP action, without introducing additional artifacts. We find that discretization effects in gradient-flow observables are highly suppressed and less than 1% up to lattice spacings of 0.14 fm, allowing continuum physics to be extracted from coarse lattices. The quality of improvement achieved motivates the use of the FP action in future gauge theory studies. The advantages of ML-based parameterizations also highlight the possibility of realizing quantum perfect actions in lattice gauge theory.

hep-lat↗
arXiv2025

$η$, $η^\prime$ mesons from lattice QCD in fully physical conditions

We determine masses and mixing parameters of the $η$ and $M_{η^\prime}$ meson in lattice QCD. The calculations are carried out on a set of 13 ETMC gauge ensembles with $N_f=2+1+1$ (maximally) twisted-mass Clover-improved quarks. These ensemble cover four values of the lattice spacing $a=0.057\mathrm{fm},...,0.092\mathrm{fm}$ and pion masses from $140\mathrm{MeV}$ to $360\mathrm{MeV}$, including three ensembles at physical quark masses and six ensembles with $M_π<200\mathrm{MeV}$. The strange-quark contribution is treated in a mixed-action approach using Osterwalder-Seiler fermions to avoid complications due to flavor mixing in the heavy quark sector and to enable the use of the one-end trick in the computation of strange quark-disconnected diagrams. With the strange-quark mass tuned to its physical value and several ensembles having close-to-physical light-quark mass, uncertainties related to the chiral extrapolations are reduced significantly compared to earlier studies. Physical results are computed with fully controlled systematics from a combined chiral, continuum and infinite-volume extrapolation, and a full error budget is obtained from model averages over of various fit ansätze and data cuts. Our results for the masses are given by $M_η=551(16)\mathrm{MeV}$ and $M_{η^\prime}=972(20)\mathrm{MeV}$, respectively, where statistical and systematic errors have been added in quadrature. For the mixing angle and decay-constant parameters the Feldmann-Kroll-Stech scheme is employed to compute them from pseudoscalar matrix elements in the quark-flavor basis. For the mixing angle we obtain $ϕ^\mathrm{phys}=39.3(2.0)^\circ$ and our results for the decay-constant parameters are given by $f_l^\mathrm{phys}=138.6(4.4)\mathrm{MeV}$ and $f_s^\mathrm{phys}=170.7(3.3)\mathrm{MeV}$.

hep-lat↗
arXiv2025

HMC and gradient flow with machine-learned classically perfect fixed-point actions

Fixed-point (FP) lattice actions are classically perfect, i.e., they have continuum classical properties unaffected by discretization effects and are expected to have suppressed lattice artifacts at weak coupling. Therefore they provide a possible way to extract continuum physics with coarser lattices, allowing to circumvent problems with critical slowing down and topological freezing towards the continuum limit. We use machine-learning methods to parameterize a FP action for four-dimensional SU(3) gauge theory using lattice gauge-covariant convolutional neural networks. The large operator space allows us to find superior parameterizations compared to previous studies and we show how such actions can be efficiently simulated with the Hybrid Monte Carlo algorithm. Furthermore, we argue that FP lattice actions can be used to define a classically perfect gradient flow without any lattice artifacts at tree level. We present initial results for scaling of the gradient flow with the FP action.

hep-lat↗
arXiv2025

Quark and gluon momentum fractions in the pion and in the kaon

We present results on the momentum fraction carried by quarks and gluons in the pion and the kaon. We employ three gauge ensembles generated with $N_f=2+1+1$ Wilson twisted-mass clover-improved fermions with physical quark masses. We perform, for the first time, a continuum extrapolation directly at the physical pion. We find that the total momentum fraction carried by quarks is $\langle x \rangle_{q, R}^π= 0.575(79)$ and $\langle x \rangle_{q,R}^{K} = 0.683(50)$ and by gluons $\langle x \rangle_{g, R}^π=0.402(53)$ and $\langle x \rangle_{g, R}^{K}=0.422(67)$ in the pion and in the kaon, respectively, in the $\overline{\mathrm{MS}}$ scheme and at the renormalization scale of 2 GeV. Having computed both the quark and gluon contributions in the continuum limit, we verify the momentum sum, finding 0.984(89) for the pion and 1.13(11) for the kaon.

hep-lat↗
arXiv2024

Machine learning a fixed point action for SU(3) gauge theory with a gauge equivariant convolutional neural network

Fixed point lattice actions are designed to have continuum classical properties unaffected by discretization effects and reduced lattice artifacts at the quantum level. They provide a possible way to extract continuum physics with coarser lattices, thereby allowing one to circumvent problems with critical slowing down and topological freezing toward the continuum limit. A crucial ingredient for practical applications is to find an accurate and compact parametrization of a fixed point action, since many of its properties are only implicitly defined. Here we use machine learning methods to revisit the question of how to parametrize fixed point actions. In particular, we obtain a fixed point action for four-dimensional SU(3) gauge theory using convolutional neural networks with exact gauge invariance. The large operator space allows us to find superior parametrizations compared to previous studies, a necessary first step for future Monte Carlo simulations and scaling studies.

hep-lat↗
arXiv2023

The $η\rightarrow γ^* γ^*$ transition form factor and the hadronic light-by-light $η$-pole contribution to the muon $g-2$ from lattice QCD

We calculate the double-virtual $η\rightarrow γ^* γ^*$ transition form factor $\mathcal{F}_{η\to γ^* γ^*}(q_1^2,q_2^2)$ from first principles using a lattice QCD simulation with $N_f=2+1+1$ quark flavors at the physical pion mass and at one lattice spacing and volume. The kinematic range covered by our calculation is complementary to the one accessible from experiment and is relevant for the $η$-pole contribution to the hadronic light-by-light scattering in the anomalous magnetic moment $a_μ= (g-2)/2$ of the muon. From the form factor calculation we extract the partial decay width $Γ(η\rightarrow γγ) = 323(85)_\text{stat}(22)_\text{syst}$ eV and the slope parameter $b_η=1.19(36)_\text{stat}(16)_\text{syst}$ GeV${}^{-2}$. For the $η$-pole contribution to $a_μ$ we obtain $a_μ^{η-\text{pole}} = 13.2(5.2)_\text{stat}(1.3)_\text{syst} \cdot 10^{-11}$.

hep-lat↗
arXiv2023

Glueballs in $N_f=1$ QCD

We present an evaluation of the glueball spectrum for configurations produced with $N_f=1$ dynamical fermions as a function of the $m_{\rm PCAC}$ mass. We obtained masses of states that fall into the irreducible representations of the octahedral group of rotations in combination with the quantum numbers of charge conjugation $C$ and parity $P$. Due to the low signal to noise ratio, practically, we can only extract masses for the irreducible representations $R^{PC}=$ $A_1^{++}$, $E^{++}$, $T_2^{++}$ as well as $A_1^{-+}$. We make use of the Generalized Eigenvalue Problem (GEVP) with an operator basis consisting only of gluonic operators. Throughout this work we are aiming towards the identification of the effects of light dynamical quarks on the glueball spectrum and how this compares to the statistically more precise spectrum of SU(3) pure gauge theory. We used large gauge ensembles which consist of ${\sim {~\cal O}}(10 {\rm K})$ configurations. Our findings demonstrate that the low-lying spectrum of the scalar, tensor as well as pseudo-scalar glueballs receive negligible contributions from the inclusion of $N_f=1$ dynamical fermions.

hep-lat↗
arXiv2023

Fixed point actions from convolutional neural networks

Lattice gauge-equivariant convolutional neural networks (L-CNNs) can be used to form arbitrarily shaped Wilson loops and can approximate any gauge-covariant or gauge-invariant function on the lattice. Here we use L-CNNs to describe fixed point (FP) actions which are based on renormalization group transformations. FP actions are classically perfect, i.e., they have no lattice artifacts on classical gauge-field configurations satisfying the equations of motion, and therefore possess scale invariant instanton solutions. FP actions are tree-level Symanzik-improved to all orders in the lattice spacing and can produce physical predictions with very small lattice artifacts even on coarse lattices. We find that L-CNNs are much more accurate at parametrizing the FP action compared to older approaches. They may therefore provide a way to circumvent critical slowing down and topological freezing towards the continuum limit.

hep-lat↗
arXiv2023

Probing the energy-smeared R-ratio on the lattice

We present a first-principles lattice QCD investigation of the $R$-ratio between the $e^+e^-$ cross-section into hadrons and that into muons. By using the method of Ref.[1], that allows to extract smeared spectral densities from Euclidean correlators, we compute the $R$-ratio convoluted with Gaussian smearing kernels of widths of about $600$ MeV and central energies from $220$ MeV up to $2.5$ GeV. Our theoretical results are compared with the corresponding quantities obtained by smearing the KNT19 compilation [2] of $R$-ratio experimental measurements with the same kernels and, by centring the Gaussians in the region around the $ρ$-resonance peak, a tension of about three standard deviations is observed. From the phenomenological perspective, we have not included yet in our calculation QED and strong isospin-breaking corrections and this might affect the observed tension. From the methodological perspective, our calculation demonstrates that it is possible to study the $R$-ratio in Gaussian energy bins on the lattice at the level of accuracy required in order to perform precision tests of the Standard Model.

hep-lat↗
arXiv2023

Transfer matrices and temporal factorization of the Wilson fermion determinant

When lattice QCD is formulated in sectors of fixed quark numbers, the canonical fermion determinants can be expressed explicitly in terms of transfer matrices. This in turn provides a complete factorization of the fermion determinants in temporal direction. Here we present a generic overview of this factorization, apply it to Wilson-type fermions and provide explicit constructions of the transfer matrices. Possible applications of the factorization include multi-level integration schemes and the construction of improved estimators for generic $n$-point correlation functions.

hep-lat↗
arXiv2022

Lattice calculation of the R-ratio smeared with Gaussian kernel

The ratio $R(E)$ of the cross-sections for $e^+e^-\to$ hadrons and $e^+e^-\to μ^+μ^-$ is a valuable energy-dependent probe of the hadronic sector of the Standard Model. Moreover, the experimental measurements of $R(E)$ are the inputs of the dispersive calculations of the leading hadronic vacuum polarization contribution to the muon $g-2$ and these are in significant tension with direct lattice calculations and with the muon $g-2$ experiment. In this talk we discuss the results of our first-principles lattice study of $R(E)$. By using a recently proposed method for extracting smeared spectral densities from Euclidean lattice correlators, we have calculated $R(E)$ convoluted with Gaussian kernels of different widths $σ$ and central energies up to $2.5$ GeV. Our theoretical results have been compared with the KNT19 [1] compilation of experimental results smeared with the same Gaussian kernels and a tension (about three standard deviations) has been observed for $σ\sim 600$ MeV and central energies around the $ρ$-resonance peak.

hep-lat↗
arXiv2022

Pseudoscalar-pole contributions to the muon $g-2$ at the physical point

Pseudoscalar-pole diagrams are an important component of estimates of the hadronic light-by-light (HLbL) contribution to the muon $g-2$. We report on our computation of the transition form factors $\mathcal{F}_{P \rightarrow γ^* γ^*}$ for the neutral pseudoscalar mesons $P=π^0$ and $η$. The calculation is performed using twisted-mass lattice QCD with physical quark masses. On the lattice, we have access to a broad range of (space-like) photon four-momenta and therefore produce form factor data complementary to the experimentally accessible single-virtual direction, which directly leads to an estimate of the pion- and $η$-pole components of the muon $g-2$. For the pion, our result for the $g-2$ contribution in the continuum is comparable with previous lattice and data-driven determinations, with combined relative uncertainties below $10\%$. For the $η$ meson, we report on a preliminary determination from a single lattice spacing.

hep-lat↗
arXiv2022

Short \& intermediate distance HVP contributions to muon g-2: SM (lattice) prediction versus $e^+e^-$ annihilation data

We present new lattice results of the ETM Collaboration, obtained from extensive simulations of lattice QCD with dynamical up, down, strange and charm quarks at physical mass values, different volumes and lattice spacings, concerning the SM prediction for the so-called intermediate window (W) and short-distance (SD) contributions to the leading order hadronic vacuum polarization (LO-HVP) term of the muon anomalous magnetic moment, $a_μ$. Results for $a_μ^{\rm LO-HVP,W}$ and $a_μ^{\rm LO-HVP,SD}$, besides representing a step forward to a complete lattice computation of $a_μ^{\rm LO-HVP}$ and a useful benchmark among lattice groups, are compared here with their dispersive counterparts based on experimental data for $e^+e^-$ into hadrons. The comparison confirms the tension in $a_μ^{\rm LO-HVP,W}$, already noted in 2020 by the BMW Collaboration, while showing no tension in $a_μ^{\rm LO-HVP,SD}$.

hep-ph↗
arXiv2022

Disconnected contribution to the LO HVP term of muon g-2 from ETMC

We present a lattice determination of the disconnected contributions to the leading-order hadronic vacuum polarization (HVP) to the muon anomalous magnetic moment in the so-called short and intermediate time-distance windows. We employ gauge ensembles produced by the Extended Twisted Mass Collaboration (ETMC) with $N_f = 2 + 1 + 1$ flavours of Wilson twisted-mass clover-improved quarks with masses approximately tuned to their physical value. We take the continuum limit employing three lattice spacings at about 0.08, 0.07 and 0.06 fm.

hep-lat↗
arXiv2022

Pion-pole contribution to HLbL from twisted mass lattice QCD at the physical point

We report on our computation of the pion transition form factor ${\cal F}_{P\rightarrow γ^*γ^*}$ from twisted mass lattice QCD in order to determine the numerically dominant light pseudoscalar pole contribution in the hadronic light-by-light scattering contribution to the anomalous magnetic moment of the muon $a_μ=(g-2)_μ$. The pion transition form factor is computed directly at the physical point. We present first results for our estimate of the pion-pole contribution with kinematic setup for the pion at rest.

hep-lat↗