Determining the NJL Coupling and AMM in Magnetized QCD Matter via Machine Learning
Zigeng Ding, Fan Lin, Xinyang Wang · 2026
Read Determining the NJL Coupling and AMM in Magnetized QCD Matter via Machine Learning by Zigeng Ding, Fan Lin, Xinyang Wang (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-29269.txt.
Introduction
Ding and co-authors use machine learning to pin down parameters of magnetized QCD matter. They work within the Nambu-Jona-Lasinio model. Two field-dependent quantities are unknown: the coupling and the anomalous magnetic moment ratio. Lattice QCD condensate data serve as ground truth. The NJL gap equation sits inside their network as a physics module. The fitted functions reproduce inverse magnetic catalysis. Both quantities fall smoothly as the magnetic field grows. The work links effective models with lattice data.
Catalog record
| CREATOR | Zigeng Ding, Fan Lin, Xinyang Wang |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 31264 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.29269
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Determining the NJL Coupling and AMM in Magnetized QCD Matter via Machine Learning Zigeng Ding, Fan Lin, Xinyang Wang Abstract In this study, we investigate the phase structure of magnetized QCD matter by determining the field-dependent parameters of the Nambu-Jona-Lasinio (NJL) model through a physics-informed machine learning framework. Specifically, we focus on extracting the optimal functional forms for the running coupling constant G ( e B ) G(eB) and the quark anomalous magnetic moment (AMM) ratio v 2 ( e B ) v_{2}(eB) , utilizing lattice QCD-computed quark condensate data as the “ground truth”. By embedding the NJL gap equation as a differentiable physics-constrained module, our neural network pipeline identifies continuous parameter functions that accurately reproduce the inverse magnetic catalysis (IMC) effect. Our results demonstrate that the magnetic field smoothly suppresses both G G and v 2 v_{2} . This approach not only bridges the gap between effective models and lattice data but also provides new microscopic insights into the response of the QCD vacuum to strong magnetic fields. I Introduction Ultra-strong magnetic fields arise in a variety of extreme environments, ranging from the transient fields produced in relativistic heavy-ion collisions to the enormous surface fields of magnetars. Such intense magnetic backgrounds provide a unique laboratory for exploring strongly interacting matter, as they can significantly modify the structure of the quantum chromodynamics (QCD) vacuum and influence the properties of hadronic and quark matter. Understanding how QCD behaves under these conditions has therefore become an important problem in nuclear and high-energy physics. In particular, the phase structure of QCD matter in strong magnetic fields remains a central puzzle, closely related to fundamental issues such as chiral symmetry breaking and modifications to the QCD phase diagram. Early theoretical work predicted that the chiral critical temperature would rise with increasing magnetic field—a phenomenon known as magnetic catalysis (MC) 30 ; 31 ; 24 ; 49 . However, subsequent lattice QCD studies revealed a strikingly different behavior around the transition region: the critical temperature for chiral symmetry restoration decreases with the magnetic field, an effect termed inverse magnetic catalysis (IMC) 3 ; 4 ; 12 . This counterintuitive phenomenon has attracted significant theoretical interest, prompting numerous investigations into its underlying physical origin, which remains incompletely understood. The Nambu–Jona-Lasinio (NJL) model 43 ; 44 , as a successful low-energy effective theory, provides a powerful and tractable framework for describing chiral symmetry breaking and meson physics. In the conventional treatment, the model’s coupling constant is held fixed, which invariably leads to magnetic catalysis 49 . Recent studies have proposed a more realistic approach by promoting the four-fermion coupling G to a magnetic-field-dependent quantity G(eB) . By calibrating G(eB) against lattice QCD data, the IMC effect can be naturally reproduced 15 ; 18 ; 36 ; 48 ; 1 ; 39 ; 41 ; 40 . This approach reflects a physically motivated picture: strong magnetic fields inevitably modify non-perturbative QCD dynamics and hadron structure. While first-principle calculations in this regime are challenging, effective models such as the NJL model offer a natural setting in which the effective coupling can “run” with the field strength, thereby encoding vacuum polarization and other non-perturbative responses to the magnetic field. Furthermore, the quark anomalous magnetic moment (AMM), another manifestation of chiral symmetry breaking, has attracted considerable attention 20 ; 22 ; 21 ; 5 ; 8 ; 54 ; 38 ; 42 ; 10 ; 23 ; 50 ; 11 ; 19 ; 17 ; 9 ; 16 . Previous studies show that the quark AMM suppresses the chiral condensate in both zero and finite-temperature regimes. Additionally, the quark AMM is an intrinsic physical parameter that becomes non-negligible in the presence of strong magnetic fields. Although the microscopic origin of the AMM is a complex problem, recent studies—borrowing from QED estimates where the lepton AMM is proportional to the lepton mass 34 —favor a magnetic-field-dependent form proportional to the square of the chiral condensate, \sigma^{2} 35 ; 28 ; 53 ; 29 . In most existing works, the proportionality ratio is treated as a constant, with field dependence entering only through the chiral condensate. More generally, however, this ratio may also depend on the magnetic field and should be taken into account. In recent years, machine learning, particularly neural networks, has demonstrated immense potential for discovering physical laws from high-dimensional data and constructing constitutive relations 47 ; 33 ; 6 ; 32 ; 2 ; 13 ; 45 ; 46 ; 14 ; 25 ; 56 ; 37 ; 52 ; 26 ; 51 ; 27 . It offers a new paradigm to bridge the gap between lattice QCD data and effective models. The core idea of this work is to employ machine learning as a powerful tool for inverse engineering, using lattice QCD computed magnetic-field-dependent quark condensate data as the “ground truth” and a parameterized (field-dependent) NJL model as the “interpretable function”. We train a neural network to directly “learn” the optimal functional forms for the evolving coupling constant G(eB) and AMM ratio v_{2}(eB) which is defined by \kappa(eB)=v_{2}(eB)\sigma^{2} . This paper aims to develop and apply this framework. We first construct a lattice-improved NJL model incorporating a magnetic-field-dependent coupling and AMM. Subsequently, we design a supervised learning pipeline: using the forward NJL model calculation as a differentiable physics-constrained module and targeting lattice QCD data, we employ automatic differentiation and gradient descent optimization to invert for a set of continuous and flexible G(eB) and v_{2}(eB) functions that accurately reproduce the lattice data. This approach avoids prior bias arising from pre-specified parameter forms and can reveal complex evolutionary patterns difficult to detect with traditional fitting methods. Ultimately, the learned parameter evolution will provide new microscopic insights into the response of the QCD vacuum to magnetic fields and yield a more precise and reliable parameterized effective model for studying the QCD phase diagram and related phenomena. The paper is structured as follows: Section II reviews the theoretical framework of the NJL model in a magnetic field and the introduction of running parameters. Section III details our machine learning methodology and model architecture. Section 2 presents the results for G(eB) and v_{2}(eB) learned from lattice data, along with physical analysis. Section V provides a summary and outlook for future work. II Theoretical Framework: NJL Model with Field-Dependent Parameters The Lagrangian density of the two-flavor Nambu–Jona-Lasinio (NJL) model in the presence of a background magnetic field, including the contribution from the anomalous magnetic moment (AMM) of quarks, is given by 17 \mathcal{L}=\bar{\psi}\left(i\gamma^{\mu}D_{\mu}-\hat{m}+\frac{1}{2}q_{f}\kappa F_{\mu\nu}\sigma^{\mu\nu}\right)\psi+G\left[(\bar{\psi}\psi)^{2}+(\bar{\psi}i\gamma^{5}\bm{\tau}\psi)^{2}\right], (1) where \hat{m}=\mathrm{diag}(m_{u},m_{d}) is the current quark mass matrix, q_{f}=\mathrm{diag}(2e/3,-e/3) denotes the quark charge matrix, and \kappa=\mathrm{diag}(\kappa_{u},\kappa_{d}) represents the AMM matrix. Here e denotes the absolute value of the electron charge. The covariant derivative D_{\mu}=\partial_{\mu}-iq_{f}A_{\mu}, (2) incorporates the coupling between quarks and the external electromagnetic field. To determine the dynamical quark mass, we employ the mean-field approximation. In this approximation, the four-fermion interaction terms are linearized as (\bar{\psi}\psi)^{2}=2\langle\bar{\psi}\psi\rangle\,\bar{\psi}\psi-\langle\bar{\psi}\psi\rangle^{2},\qquad(\bar{\psi}i\gamma^{5}\bm{\tau}\psi)^{2}=2\langle\bar{\psi}i\gamma^{5}\bm{\tau}\psi\rangle\,\bar{\psi}i\gamma^{5}\bm{\tau}\psi-\langle\bar{\psi}i\gamma^{5}\bm{\tau}\psi\rangle^{2}, (3) where \langle\bar{\psi}\psi\rangle and \langle\bar{\psi}i\gamma^{5}\bm{\tau}\psi\rangle are the chiral condensate and the pseudoscalar condensate, respectively. In the absence of an axial chemical potential \mu_{5} , the pseudoscalar condensate is assumed to vanish. The Lagrangian density then reduces to \displaystyle\mathcal{L} \displaystyle=\bar{\psi}\left(i\gamma^{\mu}D_{\mu}-\hat{m}+\frac{1}{2}q_{f}\kappa F_{\mu\nu}\sigma^{\mu\nu}+2G\langle\bar{\psi}\psi\rangle\right)\psi-G\langle\bar{\psi}\psi\rangle^{2} (4) \displaystyle=\bar{\psi}\left(i\gamma^{\mu}D_{\mu}-\hat{M}+\frac{1}{2}q_{f}\kappa F_{\mu\nu}\sigma^{\mu\nu}\right)\psi-\frac{(\hat{M}-\hat{m})^{2}}{4G}, where the constituent quark mass matrix