pith. machine review for the scientific record. sign in

arxiv: 2512.06044 · v3 · submitted 2025-12-05 · ✦ hep-lat · hep-th

Recognition: 2 theorem links

· Lean Theorem

HoloNet: Toward a Unified Einstein-Maxwell-Dilaton Framework of QCD

Hong-An Zeng , Lingxiao Wang , Mei Huang
Authors on Pith no claims yet

Pith reviewed 2026-05-17 01:45 UTC · model grok-4.3

classification ✦ hep-lat hep-th
keywords holographic QCDEinstein-Maxwell-Dilatonlattice QCDneural networksQCD phase diagramcritical end pointfinite density
0
0 comments X

The pith

A neural network learns the holographic metric and coupling from zero-density lattice QCD data and then extends the description to finite density to map the phase diagram.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper proposes HoloNet as a neural-network method that takes 2+1-flavor lattice QCD thermodynamics at vanishing chemical potential and learns the bulk metric profile A(z) together with the gauge-dilaton coupling f(z) without presupposing any particular functional form. These learned functions are inserted into the Einstein-Maxwell-Dilaton field equations, where they reproduce the lattice equation of state and baryon-number fluctuations. Because the construction is data-driven, the same functions can be used at nonzero chemical potential, where direct lattice simulations suffer from the sign problem, thereby providing a route to the QCD phase diagram and an estimate for the critical end point. The potential V(φ) and coupling reconstructed this way also match those obtained by standard holographic renormalization.

Core claim

HoloNet learns the metric profile A(z) and the gauge-dilaton coupling f(z) directly from lattice QCD data at μ=0. These functions are substituted into the Einstein-Maxwell-Dilaton equations to recover the lattice thermodynamics at zero density. The same embedding then yields predictions at finite density, including the location of the critical end point, while the derived potential and coupling remain consistent with holographic renormalization results.

What carries the argument

HoloNet, the neural-network pipeline that extracts the metric profile A(z) and coupling f(z) from zero-density lattice data and inserts them into the Einstein-Maxwell-Dilaton equations.

If this is right

  • The trained model reproduces the lattice equation of state and baryon number fluctuations at zero density with high fidelity.
  • The same functions allow direct extension to finite chemical potential, where lattice methods are limited by the sign problem.
  • The framework produces a map of the QCD phase diagram and an estimate for the location of the critical end point.
  • The reconstructed potential V(φ) and coupling f(φ) agree quantitatively with results from holographic renormalization.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The method could be retrained on additional lattice observables such as susceptibilities or correlation functions to further constrain the bulk geometry.
  • Similar data-driven extraction might be applied to other gauge theories where holographic duals are conjectured but functional forms are unknown.
  • Direct comparison with small-μ lattice data could reveal whether the learned functions require explicit density dependence.

Load-bearing premise

The metric profile and coupling learned exclusively from zero-density lattice data remain the correct functions to use in the Einstein-Maxwell-Dilaton equations once a nonzero chemical potential is introduced.

What would settle it

If independent calculations or future lattice results at small but nonzero chemical potential show that the predicted critical end point location or equation of state deviates significantly from the HoloNet extrapolation, the assumption that no density-dependent corrections are needed would be falsified.

Figures

Figures reproduced from arXiv: 2512.06044 by Hong-An Zeng, Lingxiao Wang, Mei Huang.

Figure 1
Figure 1. Figure 1: FIG. 1 [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2. The graphs of the [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3. The figures display the speed of sound and specific heat, [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4. EoS comparison between neural network results and [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: FIG. 5. Baryon number susceptibility comparison between [PITH_FULL_IMAGE:figures/full_fig_p007_5.png] view at source ↗
Figure 7
Figure 7. Figure 7: FIG. 7. Coupling function [PITH_FULL_IMAGE:figures/full_fig_p007_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: FIG. 8. The phase diagram is shown, where the red solid [PITH_FULL_IMAGE:figures/full_fig_p008_8.png] view at source ↗
read the original abstract

We propose HoloNet, a neural-network framework that unifies lattice QCD(LQCD) thermodynamics and holographic Einstein-Maxwell-Dilaton (EMD) theory within a data-to-holography pipeline. Instead of assuming specific functional forms, HoloNet learns the metric profile $A(z)$ and the gauge-dilaton coupling $f(z)$ directly from 2+1-flavor LQCD data at $\mu=0$. These learned functions are embedded into the EMD equations, enabling the model to reproduce the lattice equation of state and baryon number fluctuations with high fidelity. Once trained, HoloNet provides a fully data-driven holographic description of QCD that extends naturally to finite density, allowing us to map the phase diagram and estimate the location of the critical end point (CEP). The reconstructed potential $V(\phi)$ and coupling $f(\phi)$ agree quantitatively with those obtained from holographic renormalization, demonstrating that HoloNet can consistently bridge different holographic models.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 2 minor

Summary. The manuscript introduces HoloNet, a neural-network framework that learns the metric profile A(z) and gauge-dilaton coupling f(z) directly from 2+1-flavor lattice QCD data at μ=0 without assuming specific functional forms. These functions are embedded into the Einstein-Maxwell-Dilaton equations to reproduce the lattice equation of state and baryon number fluctuations at zero density with claimed high fidelity. The trained model is then extended to finite chemical potential to map the QCD phase diagram and estimate the location of the critical end point, with reconstructed V(φ) and f(φ) shown to agree quantitatively with holographic renormalization results.

Significance. If the central extrapolation holds, the work could provide a useful data-driven bridge between lattice QCD and holographic EMD models, enabling finite-density predictions where direct lattice methods face the sign problem. The avoidance of ad-hoc functional forms for potentials and the quantitative μ=0 reproduction are positive features. However, the significance is tempered by the lack of independent validation for the finite-μ regime, which risks making the CEP estimate an uncontrolled continuation of the zero-density fit rather than a robust prediction.

major comments (3)
  1. [Finite-density extension and results] The central finite-density claim (phase diagram mapping and CEP location) rests on inserting A(z) and f(z) learned exclusively from μ=0 data into the full EMD equations with nonzero chemical potential boundary conditions. No test, constraint, or justification is provided that these functions acquire no density-dependent corrections; this assumption is load-bearing and unverified.
  2. [Abstract and μ=0 reproduction section] The abstract states high-fidelity reproduction of the equation of state and fluctuations at μ=0, yet supplies no quantitative error bars, cross-validation statistics, or explicit checks on whether the zero-density functions remain valid when μ is switched on. This weakens confidence in the subsequent extrapolation.
  3. [Potential reconstruction] The reconstruction of V(φ) and f(φ) is reported to agree quantitatively with holographic renormalization, but it is unclear from the training procedure how the neural-network outputs for A(z) and f(z) are converted to these potentials and whether this agreement independently supports the finite-μ predictions.
minor comments (2)
  1. Clarify the neural-network architecture with an explicit diagram or layer equations to aid reproducibility.
  2. [Training procedure] Specify the exact lattice data sets employed (temperature range, number of configurations, action details) in the training description.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the careful reading and constructive comments on our manuscript. We address each major comment below and have revised the manuscript to strengthen the presentation and clarify key assumptions.

read point-by-point responses

  1. Referee: [Finite-density extension and results] The central finite-density claim (phase diagram mapping and CEP location) rests on inserting A(z) and f(z) learned exclusively from μ=0 data into the full EMD equations with nonzero chemical potential boundary conditions. No test, constraint, or justification is provided that these functions acquire no density-dependent corrections; this assumption is load-bearing and unverified.

    Authors: We appreciate the referee highlighting this foundational assumption. Within the EMD holographic framework, A(z) and f(z) encode the dual description of the QCD vacuum structure as determined by μ=0 lattice data; finite-density effects enter solely through the Maxwell boundary conditions while the bulk functions remain fixed, consistent with standard practice in holographic QCD models. Direct validation at finite μ is precluded by the sign problem, but we have added a new discussion paragraph explaining this rationale, referencing analogous extrapolations in the EMD literature, and explicitly noting the assumption as a model limitation. revision: yes

  2. Referee: [Abstract and μ=0 reproduction section] The abstract states high-fidelity reproduction of the equation of state and fluctuations at μ=0, yet supplies no quantitative error bars, cross-validation statistics, or explicit checks on whether the zero-density functions remain valid when μ is switched on. This weakens confidence in the subsequent extrapolation.

    Authors: We agree that quantitative metrics improve transparency. The revised manuscript now reports explicit reproduction accuracies (typically within 1–3% for thermodynamic quantities and susceptibilities), includes error bands on all μ=0 comparison figures, and adds a validation subsection with cross-validation results across lattice ensembles. We also demonstrate that the learned functions produce stable results for small nonzero μ before the full extrapolation. revision: yes

  3. Referee: [Potential reconstruction] The reconstruction of V(φ) and f(φ) is reported to agree quantitatively with holographic renormalization, but it is unclear from the training procedure how the neural-network outputs for A(z) and f(z) are converted to these potentials and whether this agreement independently supports the finite-μ predictions.

    Authors: We thank the referee for noting the need for procedural clarity. We have expanded the methods section and added an appendix that details the conversion: the NN outputs A(z) and f(z) are inserted into the EMD equations and solved via holographic renormalization to obtain V(φ) and the φ-dependent coupling. This quantitative agreement validates that HoloNet recovers standard holographic potentials from data without ad-hoc assumptions. While it does not constitute independent finite-μ validation, it confirms internal consistency of the zero-density training and thereby lends support to the overall framework. revision: partial

Circularity Check

0 steps flagged

No significant circularity in derivation chain

full rationale

The paper trains HoloNet exclusively on μ=0 LQCD data to determine A(z) and f(z), then inserts these functions into the EMD equations. Reproduction of zero-density thermodynamics and fluctuations occurs by construction as a fit. However, the finite-density extension (phase diagram, CEP location) is not equivalent to the inputs by construction; it rests on the explicit assumption that A(z) and f(z) carry no additional μ dependence. This assumption is externally falsifiable by future nonzero-density lattice data or other holographic models and does not reduce any claimed prediction to a renaming or re-use of the training set. No self-definitional equations, load-bearing self-citations, or uniqueness theorems imported from the same authors appear in the abstract or described pipeline. The derivation therefore remains self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The central claim rests on the assumption that a single set of bulk functions learned at zero density can be used unchanged at finite density inside the standard EMD action, plus the usual holographic dictionary that maps bulk fields to boundary QCD operators.

free parameters (1)
  • Neural-network weights and architecture hyperparameters
    The network that parametrizes A(z) and f(z) contains a large number of adjustable parameters fitted to lattice data.
axioms (1)
  • domain assumption The Einstein-Maxwell-Dilaton action in five dimensions correctly encodes the thermodynamics of 2+1-flavor QCD.
    Invoked when the learned A(z) and f(z) are inserted into the EMD equations to generate finite-density results.

pith-pipeline@v0.9.0 · 5471 in / 1453 out tokens · 29219 ms · 2026-05-17T01:45:15.801230+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

  • IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel echoes
    ?
    echoes

    ECHOES: this paper passage has the same mathematical shape or conceptual pattern as the Recognition theorem, but is not a direct formal dependency.

    HoloNet learns the metric profile A(z) and the gauge-dilaton coupling f(z) directly from 2+1-flavor LQCD data at μ=0. These learned functions are embedded into the EMD equations... reconstructed potential V(ϕ) and coupling f(ϕ) agree quantitatively with those obtained from holographic renormalization.

  • IndisputableMonolith/Foundation/AlphaCoordinateFixation.lean costAlphaLog_high_calibrated_iff echoes
    ?
    echoes

    ECHOES: this paper passage has the same mathematical shape or conceptual pattern as the Recognition theorem, but is not a direct formal dependency.

    V(ϕ) = −12 cosh(c1 ϕ) + c2 ϕ², f(ϕ) = c3 sech(c4(ϕ + c5)³)

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

88 extracted references · 88 canonical work pages · 3 internal anchors

  1. [1]

    The sub-networks dy- namically generate the layer-wise values of A(z) and f (z)

    Self-adaptive optimization . The sub-networks dy- namically generate the layer-wise values of A(z) and f (z). Previous approaches [64] treat these values as direct pa- rameters of the global network, reconstructing the geom- etry through their optimization. By contrast, our model computes them on the fly through explicit neural net- works. Although seeming...

    work page

  2. [2]

    21002, a4 = 0

    062182, a3 = 0 . 21002, a4 = 0 . 020314, f1 = 0 . 31197, f2 = 0. 079030, f3 = 0. 34070. The analytic and neural-network results agree very well, with the analytic curves shown in black in Figs. 2, 6, 7, and 8. In the temperature range 130–400 MeV, the maximal relative discrepancies between the neural- network reconstruction and the LQCD data are of order ...

    work page

  3. [3]

    061371, c5 = 0. 35482. 8 C. Extrapolation to CEP We have also computed the locations of the CEP, the first-order phase transition line, and the crossover in the phase diagram in Fig. 8. Overall, our results are some- what higher in temperature compared with most previous studies. Fortunately, these results fall within the region represented by our data, ma...

    work page 2021

  4. [4]

    The phase structure of qcd

    Christian Schmidt and Sayantan Sharma. The phase structure of qcd. Journal of Physics G: Nuclear and Particle Physics, 44(10):104002, 2017

    work page 2017

  5. [5]

    Higher moments of net proton multiplicity distributions at rhic

    MM Aggarwal, Z Ahammed, A V Alakhverdyants, I Alekseev, J Alford, BD Anderson, D Arkhipkin, GS Averichev, J Balewski, LS Barnby, et al. Higher moments of net proton multiplicity distributions at rhic. Physical review letters, 105(2):022302, 2010

    work page 2010

  6. [6]

    An Experimental Exploration of the QCD Phase Diagram: The Search for the Critical Point and the Onset of De-confinement

    MM Aggarwal, Z Ahammed, A V Alakhverdyants, I Alek- seev, BD Anderson, D Arkhipkin, GS Averichev, J Balewski, LS Barnby, S Baumgart, et al. An experi- mental exploration of the qcd phase diagram: the search for the critical point and the onset of de-confinement. arXiv preprint arXiv:1007.2613, 2010

    work page internal anchor Pith review Pith/arXiv arXiv 2010

  7. [7]

    Energy dependence of moments of net-proton multiplicity distributions at rhic

    L Adamczyk, JK Adkins, G Agakishiev, MM Aggarwal, Z Ahammed, I Alekseev, J Alford, CD Anson, A Aparin, D Arkhipkin, et al. Energy dependence of moments of net-proton multiplicity distributions at rhic. Physical Review Letters, 112(3):032302, 2014

    work page 2014

  8. [8]

    Search for the qcd critical point with fluctuations of conserved quantities in relativistic heavy-ion collisions at rhic: an overview

    Xiaofeng Luo and Nu Xu. Search for the qcd critical point with fluctuations of conserved quantities in relativistic heavy-ion collisions at rhic: an overview. Nuclear Science and Techniques, 28(8):112, 2017

    work page 2017

  9. [9]

    Nonmonotonic energy dependence of net-proton number fluctuations

    Jaroslav Adam, L Adamczyk, JR Adams, JK Adkins, G Agakishiev, MM Aggarwal, Z Ahammed, I Alekseev, DM Anderson, A Aparin, et al. Nonmonotonic energy dependence of net-proton number fluctuations. Physical review letters, 126(9):092301, 2021

    work page 2021

  10. [10]

    Higher-order cu- mulants and correlation functions of proton multiplicity distributions in s nn= 3 gev au+ au collisions at the rhic star experiment

    Mohamed Samy Abdallah, BE Aboona, Jaroslav Adam, Leszek Adamczyk, Joseph R Adams, J Kevin Ad- kins, Ishu Aggarwal, Madan Mohan Aggarwal, Zubayer Ahammed, Derek M Anderson, et al. Higher-order cu- mulants and correlation functions of proton multiplicity distributions in s nn= 3 gev au+ au collisions at the rhic star experiment. Physical Review C, 107(2):02...

    work page 2023

  11. [11]

    The QCD equation of state from the lattice

    Owe Philipsen. The QCD equation of state from the lattice. Prog. Part. Nucl. Phys., 70:55–107, 2013

    work page 2013

  12. [12]

    The qcd equation of state from the lat- tice

    Owe Philipsen. The qcd equation of state from the lat- tice. Progress in Particle and Nuclear Physics, 70:55–107, 2013

    work page 2013

  13. [13]

    Introductory lectures on lattice QCD at nonzero baryon number

    Gert Aarts. Introductory lectures on lattice QCD at nonzero baryon number. J. Phys. Conf. Ser., 706(2):022004, 2016

    work page 2016

  14. [14]

    Finite-density lattice QCD and sign problem: Current status and open problems

    Keitaro Nagata. Finite-density lattice QCD and sign problem: Current status and open problems. Prog. Part. Nucl. Phys., 127:103991, 2022

    work page 2022

  15. [15]

    Qcd phase structure at finite temperature and density

    Wei-jie Fu, Jan M Pawlowski, and Fabian Rennecke. Qcd phase structure at finite temperature and density. Physical Review D, 101(5):054032, 2020

    work page 2020

  16. [16]

    Func- tional renormalization group study of the quark-meson model with ω meson

    Hui Zhang, Defu Hou, Toru Kojo, and Bin Qin. Func- tional renormalization group study of the quark-meson model with ω meson. Physical Review D, 96(11):114029, 2017

    work page 2017

  17. [17]

    Qcd phase transitions via a re- fined truncation of dyson-schwinger equations

    Fei Gao and Yu-xin Liu. Qcd phase transitions via a re- fined truncation of dyson-schwinger equations. Physical Review D, 94(7):076009, 2016

    work page 2016

  18. [18]

    Phase diagram and critical end point for strongly interacting quarks

    Si-xue Qin, Lei Chang, Huan Chen, Yu-xin Liu, and Craig D Roberts. Phase diagram and critical end point for strongly interacting quarks. Physical Review Letters, 106(17):172301, 2011

    work page 2011

  19. [19]

    Locate qcd critical end point in a con- tinuum model study

    Chao Shi, Yong-long Wang, Yu Jiang, Zhu-fang Cui, and Hong-Shi Zong. Locate qcd critical end point in a con- tinuum model study. Journal of High Energy Physics, 2014(7):1–10, 2014

    work page 2014

  20. [20]

    Phase structure of three and four flavor qcd

    Christian S Fischer, Jan Luecker, and Christian A Welzbacher. Phase structure of three and four flavor qcd. Physical Review D, 90(3):034022, 2014. 10

    work page 2014

  21. [21]

    Phase diagram of qcd

    MA Halasz, AD Jackson, RE Shrock, Misha A Stephanov, and JJM Verbaarschot. Phase diagram of qcd. Physical Review D, 58(9):096007, 1998

    work page 1998

  22. [22]

    Dynamical model of elementary particles based on an analogy with superconductivity

    Yoichiro Nambu and Giovanni Jona-Lasinio. Dynamical model of elementary particles based on an analogy with superconductivity. i. Physical review, 122(1):345, 1961

    work page 1961

  23. [23]

    Splitting of chiral and deconfinement phase transitions induced by rotation

    Fei Sun, Kun Xu, and Mei Huang. Splitting of chiral and deconfinement phase transitions induced by rotation. Physical Review D, 108(9):096007, 2023

    work page 2023

  24. [24]

    The kurtosis of net baryon number fluctuations from a realis- tic polyakov–nambu–jona-lasinio model along the exper- imental freeze-out line

    Zhibin Li, Kun Xu, Xinyang Wang, and Mei Huang. The kurtosis of net baryon number fluctuations from a realis- tic polyakov–nambu–jona-lasinio model along the exper- imental freeze-out line. The European Physical Journal C, 79(3):245, 2019

    work page 2019

  25. [25]

    Qcd phase diagram at finite baryon and isospin chemical potentials

    Takahiro Sasaki, Yuji Sakai, Hiroaki Kouno, and Masanobu Yahiro. Qcd phase diagram at finite baryon and isospin chemical potentials. Physical Review D—Particles, Fields, Gravitation, and Cosmology, 82(11):116004, 2010

    work page 2010

  26. [26]

    Quarkyonic matter and chiral symmetry breaking

    Larry McLerran, Krzysztof Redlich, and Chihiro Sasaki. Quarkyonic matter and chiral symmetry breaking. Nuclear Physics A, 824(1-4):86–100, 2009

    work page 2009

  27. [27]

    The large-n limit of superconformal field theories and supergravity

    Juan Maldacena. The large-n limit of superconformal field theories and supergravity. International journal of theoretical physics, 38(4):1113–1133, 1999

    work page 1999

  28. [28]

    Anti De Sitter Space And Holography

    Edward Witten. Anti de sitter space and holography. arXiv preprint hep-th/9802150, 1998

    work page internal anchor Pith review Pith/arXiv arXiv 1998

  29. [29]

    Gauge theory correlators from non-critical string theory

    Steven S Gubser, Igor R Klebanov, and Alexander M Polyakov. Gauge theory correlators from non-critical string theory. Physics Letters B, 428(1-2):105–114, 1998

    work page 1998

  30. [30]

    Large n field theories, string theory and gravity

    Ofer Aharony, Steven S Gubser, Juan Maldacena, Hi- rosi Ooguri, and Yaron Oz. Large n field theories, string theory and gravity. Physics Reports, 323(3-4):183–386, 2000

    work page 2000

  31. [31]

    Qcd and a holographic model of hadrons

    Joshua Erlich, Emanuel Katz, Dam T Son, and Mikhail A Stephanov. Qcd and a holographic model of hadrons. Physical review letters, 95(26):261602, 2005

    work page 2005

  32. [32]

    Linear confinement and ads/qcd

    Andreas Karch, Emanuel Katz, Dam T Son, and Mikhail A Stephanov. Linear confinement and ads/qcd. Physical Review D—Particles, Fields, Gravitation, and Cosmology, 74(1):015005, 2006

    work page 2006

  33. [33]

    Holo- graphic qcd in the veneziano limit at a finite magnetic field and chemical potential

    Umut G¨ ursoy, Matti J¨ arvinen, and Govert Nijs. Holo- graphic qcd in the veneziano limit at a finite magnetic field and chemical potential. Physical Review Letters, 120(24):242002, 2018

    work page 2018

  34. [34]

    Neu- ral ordinary differential equations for mapping the mag- netic qcd phase diagram via holography

    Rong-Gen Cai, Song He, Li Li, and Hong-An Zeng. Neu- ral ordinary differential equations for mapping the mag- netic qcd phase diagram via holography. arXiv preprint arXiv:2406.12772, 2024

    work page arXiv 2024

  35. [35]

    A holographic critical point

    Oliver DeWolfe, Steven S Gubser, and Christopher Rosen. A holographic critical point. Physical Review D—Particles, Fields, Gravitation, and Cosmology, 83(8):086005, 2011

    work page 2011

  36. [36]

    Mimicking the qcd equation of state with a dual black hole

    Steven S Gubser and Abhinav Nellore. Mimicking the qcd equation of state with a dual black hole. Physical Review D—Particles, Fields, Gravitation, and Cosmology, 78(8):086007, 2008

    work page 2008

  37. [37]

    Hot and dense quark-gluon plasma thermo- dynamics from holographic black holes

    Joaquin Grefa, Jorge Noronha, Jacquelyn Noronha- Hostler, Israel Portillo, Claudia Ratti, and Romulo Rougemont. Hot and dense quark-gluon plasma thermo- dynamics from holographic black holes. Physical Review D, 104(3):034002, 2021

    work page 2021

  38. [38]

    Gravitational waves and primordial black hole produc- tions from gluodynamics by holography

    Song He, Li Li, Zhibin Li, and Shao-Jiang Wang. Gravitational waves and primordial black hole produc- tions from gluodynamics by holography. Science China Physics, Mechanics & Astronomy, 67(4):240411, 2024

    work page 2024

  39. [39]

    Probing qcd critical point and induced gravitational wave by black hole physics

    Rong-Gen Cai, Song He, Li Li, and Yuan-Xu Wang. Probing qcd critical point and induced gravitational wave by black hole physics. Physical Review D, 106(12):L121902, 2022

    work page 2022

  40. [40]

    Holographic entanglement properties in the qcd phase diagram from einstein-maxwell-dilaton models

    Zhibin Li. Holographic entanglement properties in the qcd phase diagram from einstein-maxwell-dilaton models. Physical Review D, 110(4):046012, 2024

    work page 2024

  41. [41]

    Revisiting holographic model for thermal and dense qcd with a crit- ical point

    Qingxuan Fu, Song He, Li Li, and Zhibin Li. Revisiting holographic model for thermal and dense qcd with a crit- ical point. Journal of High Energy Physics, 2025(6):1–19, 2025

    work page 2025

  42. [42]

    A Refined Holographic QCD Model and QCD Phase Structure

    Yi Yang and Pei-Hung Yuan. A Refined Holographic QCD Model and QCD Phase Structure. JHEP, 11:149, 2014

    work page 2014

  43. [43]

    Confinement- deconfinement phase transition for heavy quarks in a soft wall holographic qcd model

    Yi Yang and Pei-Hung Yuan. Confinement- deconfinement phase transition for heavy quarks in a soft wall holographic qcd model. Journal of High Energy Physics, 2015(12):1–22, 2015

    work page 2015

  44. [44]

    Thermal entropy of a quark-antiquark pair above and below deconfine- ment from a dynamical holographic qcd model

    David Dudal and Subhash Mahapatra. Thermal entropy of a quark-antiquark pair above and below deconfine- ment from a dynamical holographic qcd model. Physical Review D, 96(12):126010, 2017

    work page 2017

  45. [45]

    Chiral and de- confining phase transitions from holographic qcd study

    Zhen Fang, Song He, and Danning Li. Chiral and de- confining phase transitions from holographic qcd study. Nuclear Physics B, 907:187–207, 2016

    work page 2016

  46. [46]

    Thermodynamics of deformed ads5 model with a pos- itive/negative quadratic correction in graviton-dilaton system

    Danning Li, Song He, Mei Huang, and Qi-Shu Yan. Thermodynamics of deformed ads5 model with a pos- itive/negative quadratic correction in graviton-dilaton system. Journal of High Energy Physics, 2011(9):1–38, 2011

    work page 2011

  47. [47]

    Criticality of qcd in a holographic qcd model with critical end point

    Xun Chen, Danning Li, and Mei Huang. Criticality of qcd in a holographic qcd model with critical end point. Chinese Physics C, 43(2):023105, 2019

    work page 2019

  48. [48]

    Gluodynamics and deconfinement phase transi- tion under rotation from holography

    Xun Chen, Lin Zhang, Danning Li, Defu Hou, and Mei Huang. Gluodynamics and deconfinement phase transi- tion under rotation from holography. Journal of High Energy Physics, 2021(7):1–28, 2021

    work page 2021

  49. [49]

    Thermodynamics of heavy quarkonium in a magnetic field background

    Jing Zhou, Xun Chen, Yan-Qing Zhao, and Jialun Ping. Thermodynamics of heavy quarkonium in a magnetic field background. Physical Review D, 102(8):086020, 2020

    work page 2020

  50. [50]

    Machine learning holographic black hole from lattice qcd equation of state

    Xun Chen and Mei Huang. Machine learning holographic black hole from lattice qcd equation of state. Physical Review D, 109(5):L051902, 2024

    work page 2024

  51. [51]

    The entanglement properties of holographic qcd model with a critical end point

    Zhibin Li, Kun Xu, and Mei Huang. The entanglement properties of holographic qcd model with a critical end point. Chinese Physics C, 45(1):013116, 2021

    work page 2021

  52. [52]

    Dynamical holographic qcd model for glueball and light meson spectra

    Danning Li and Mei Huang. Dynamical holographic qcd model for glueball and light meson spectra. Journal of High Energy Physics, 2013(11):1–51, 2013

    work page 2013

  53. [53]

    The dy- namical holographic qcd method for hadron physics and qcd matter

    Yidian Chen, Danning Li, and Mei Huang. The dy- namical holographic qcd method for hadron physics and qcd matter. Communications in Theoretical Physics, 74(9):097201, 2022

    work page 2022

  54. [54]

    Physics- driven learning for inverse problems in quantum chromo- dynamics

    Gert Aarts, Kenji Fukushima, Tetsuo Hatsuda, Andreas Ipp, Shuzhe Shi, Lingxiao Wang, and Kai Zhou. Physics- driven learning for inverse problems in quantum chromo- dynamics. Nature Reviews Physics, 7(3):154–163, 2025

    work page 2025

  55. [55]

    Ads/deep-learning made easy: sim- ple examples

    Mugeon Song, Maverick SH Oh, Yongjun Ahn, and Keun-Young Kima. Ads/deep-learning made easy: sim- ple examples. Chinese Physics C, 45(7):073111, 2021. 11

    work page 2021

  56. [56]

    Ads/deep- learning made easy ii: neural network-based approaches to holography and inverse problems

    Hyun-Sik Jeong, Hanse Kim, Keun-Young Kim, Gaya Yun, Hyeonwoo Yu, and Kwan Yun. Ads/deep- learning made easy ii: neural network-based approaches to holography and inverse problems. arXiv preprint arXiv:2511.22522, 2025

    work page arXiv 2025

  57. [57]

    Holographic reconstruction of black hole spacetime: machine learning and entanglement entropy

    Byoungjoon Ahn, Hyun-Sik Jeong, Keun-Young Kim, and Kwan Yun. Holographic reconstruction of black hole spacetime: machine learning and entanglement entropy. Journal of High Energy Physics, 2025(1):1–41, 2025

    work page 2025

  58. [58]

    Neural ordinary differential equations

    Ricky TQ Chen, Yulia Rubanova, Jesse Bettencourt, and David K Duvenaud. Neural ordinary differential equations. Advances in neural information processing systems, 31, 2018

    work page 2018

  59. [59]

    Data- driven einstein-dilaton model for pure yang-mills ther- modynamics and glueball spectrum

    Xun Chen, Yidian Chen, and Kai Zhou. Data- driven einstein-dilaton model for pure yang-mills ther- modynamics and glueball spectrum. arXiv preprint arXiv:2507.06729, 2025

    work page arXiv 2025

  60. [60]

    Machine-learning emergent spacetime from linear response in future table- top quantum gravity experiments

    Koji Hashimoto, Koshiro Matsuo, Masaki Murata, Gakuto Ogiwara, and Daichi Takeda. Machine-learning emergent spacetime from linear response in future table- top quantum gravity experiments. Machine Learning: Science and Technology, 6(1):015030, 2025

    work page 2025

  61. [61]

    Uni- fication of symmetries inside neural networks: trans- former, feedforward and neural ode

    Koji Hashimoto, Yuji Hirono, and Akiyoshi Sannai. Uni- fication of symmetries inside neural networks: trans- former, feedforward and neural ode. Machine Learning: Science and Technology, 5(2):025079, 2024

    work page 2024

  62. [62]

    Deep learning bulk spacetime from boundary optical conductivity

    Byoungjoon Ahn, Hyun-Sik Jeong, Keun-Young Kim, and Kwan Yun. Deep learning bulk spacetime from boundary optical conductivity. Journal of High Energy Physics, 2024(3):1–30, 2024

    work page 2024

  63. [63]

    Deep learning- based holography for t-linear resistivity

    Byoungjoon Ahn, Hyun-Sik Jeong, Chang-Woo Ji, Keun-Young Kim, and Kwan Yun. Deep learning- based holography for t-linear resistivity. arXiv preprint arXiv:2502.10245, 2025

    work page arXiv 2025

  64. [64]

    Neural ordinary differential equation and holographic quantum chromodynamics

    Koji Hashimoto, Hong-Ye Hu, and Yi-Zhuang You. Neural ordinary differential equation and holographic quantum chromodynamics. Mach. Learn. Sci. Tech., 2(3):035011, 2021

    work page 2021

  65. [65]

    Deep learning and ads/qcd

    Tetsuya Akutagawa, Koji Hashimoto, and Takayuki Sumimoto. Deep learning and ads/qcd. Physical Review D, 102(2):026020, 2020

    work page 2020

  66. [66]

    Neu- ral ordinary differential equation and holographic quan- tum chromodynamics

    Koji Hashimoto, Hong-Ye Hu, and Yi-Zhuang You. Neu- ral ordinary differential equation and holographic quan- tum chromodynamics. Machine Learning: Science and Technology, 2(3):035011, 2021

    work page 2021

  67. [67]

    Deep learning and holographic qcd

    Koji Hashimoto, Sotaro Sugishita, Akinori Tanaka, and Akio Tomiya. Deep learning and holographic qcd. Physical Review D, 98(10):106014, 2018

    work page 2018

  68. [68]

    Deriving the dilaton potential in improved holo- graphic qcd from the meson spectrum

    Koji Hashimoto, Keisuke Ohashi, and Takayuki Sumi- moto. Deriving the dilaton potential in improved holo- graphic qcd from the meson spectrum. Physical Review D, 105(10):106008, 2022

    work page 2022

  69. [69]

    The holo- graphic weyl anomaly

    Mark Henningson and Kostas Skenderis. The holo- graphic weyl anomaly. Journal of High Energy Physics, 1998(07):023, 1998

    work page 1998

  70. [70]

    A stress ten- sor for anti-de sitter gravity

    Vijay Balasubramanian and Per Kraus. A stress ten- sor for anti-de sitter gravity. Communications in Mathematical Physics, 208(2):413–428, 1999

    work page 1999

  71. [71]

    Holographic reconstruction of spacetime and renormalization in the ads/cft correspondence

    Sebastian De Haro, Kostas Skenderis, and Sergey N Solodukhin. Holographic reconstruction of spacetime and renormalization in the ads/cft correspondence. Communications in Mathematical Physics, 217(3):595– 622, 2001

    work page 2001

  72. [72]

    In which, A and potential function V are mutually de- termined and are not independent variables

    work page

  73. [73]

    Machine learning holographic black hole from lattice QCD equation of state

    Xun Chen and Mei Huang. Machine learning holographic black hole from lattice QCD equation of state. Phys. Rev. D, 109(5):L051902, 2024

    work page 2024

  74. [74]

    Bayesian inference of the critical end point in a (2+ 1)-flavor system from holographic qcd

    Liqiang Zhu, Xun Chen, Kai Zhou, Hanzhong Zhang, and Mei Huang. Bayesian inference of the critical end point in a (2+ 1)-flavor system from holographic qcd. Physical Review D, 112(2):026019, 2025

    work page 2025

  75. [75]

    Spectra of glueballs and oddballs and the equation of state from holographic qcd

    Lin Zhang, Chutian Chen, Yidian Chen, and Mei Huang. Spectra of glueballs and oddballs and the equation of state from holographic qcd. Physical Review D, 105(2):026020, 2022

    work page 2022

  76. [76]

    Katz, Stefan Krieg, and Kalman K

    Szabocls Borsanyi, Zoltan Fodor, Christian Hoelbling, Sandor D. Katz, Stefan Krieg, and Kalman K. Szabo. Full result for the QCD equation of state with 2+1 fla- vors. Phys. Lett. B, 730:99–104, 2014

    work page 2014

  77. [77]

    Training behavior of deep neural network in frequency domain

    Zhi-Qin John Xu, Yaoyu Zhang, and Yanyang Xiao. Training behavior of deep neural network in frequency domain. arXiv e-prints, pages arXiv–1807, 2018

    work page 2018

  78. [78]

    Equation of state in (2+ 1)-flavor qcd

    Alexei Bazavov, Tanmoy Bhattacharya, Carleton De- Tar, H-T Ding, Steven Gottlieb, Rajan Gupta, P Hegde, UM Heller, Frithjof Karsch, Edwin Laermann, et al. Equation of state in (2+ 1)-flavor qcd. Physical Review D, 90(9):094503, 2014

    work page 2014

  79. [79]

    Fluctuations of conserved charges at finite temperature from lattice qcd

    Szabolcs Borsanyi, Zoltan Fodor, Sandor D Katz, Stefan Krieg, Claudia Ratti, and Kalman Szabo. Fluctuations of conserved charges at finite temperature from lattice qcd. Journal of High Energy Physics, 2012(1):1–15, 2012

    work page 2012

  80. [80]

    Interpretable Machine Learning for Science with PySR and SymbolicRegression.jl

    Miles Cranmer. Interpretable machine learning for sci- ence with pysr and symbolicregression. jl. arXiv preprint arXiv:2305.01582, 2023

    work page internal anchor Pith review Pith/arXiv arXiv 2023

Showing first 80 references.