QCD-Like Holography and Phenomenological Limits

QCD-like holographic models can organize confinement, chiral symmetry breaking, thermal crossover, spectra, and transport in a common geometric calculation. Their inference ceiling depends on construction. A top-down brane model has a controlled dictionary for a large-$N$, strong-coupling theory that is not real-world QCD; a bottom-up model can be calibrated closer to QCD data but lacks a unique microscopic derivation. Neither status permits a good fit to be called identification with the quark–gluon plasma.

Required background. Weak-Coupling Plasma Transport Coefficients supplies the perturbative regime and its uncertainties. Kubo Formulae and Horizon Response supplies the holographic transport observable.

Helpful background. Static Sources, Center Symmetry, and String Breaking in QCD supplies QCD order parameters and dynamical-quark qualifications. Evidence and Model Discrimination at Quantum Criticality supplies cross-model inference rules. Hydrodynamization and the Kinetic-to-Hydrodynamic Map supplies the collision-model interface.

Evidence cutoff: 25 July 2026. Phenomenological comparisons and uncertainty statements are restricted to sources available by this date.

Top-down and bottom-up questions differ

The Sakai–Sugimoto construction embeds probe D8 and anti-D8 branes in the near-horizon D4 geometry. Their connected embedding geometrizes chiral-symmetry breaking and yields towers of mesons with couplings fixed by a small set of parameters Sakai and Sugimoto 2005. Its control requires large color number and large ‘t Hooft coupling, while the compactification carries Kaluza–Klein modes that do not parametrically decouple in the QCD-like regime. It is top-down for its specified string theory, not a controlled dual of three-color QCD.

A bottom-up Einstein–dilaton model may instead use

$$S=\frac{1}{16\pi G_5}\int d^5x\sqrt{\lvert g\rvert} \left[R+\frac{4}{3}\frac{(\partial\lambda)^2}{\lambda^2}+V(\lambda)\right].$$

The ultraviolet form of $V(\lambda)$ can encode perturbative beta-function information and its infrared form can be chosen to produce confinement. Improved holographic QCD develops this strategy systematically Gürsoy and Kiritsis 2008. The interpolating potential and flavor sector remain modeling choices; changing them can preserve an equation-of-state fit while changing transport.

These two constructions answer different questions. The first demonstrates mechanisms within a derived large-$N$ theory. The second tests whether a compact gravitational parameterization can reproduce and predict selected dimensionless QCD observables.

Calibrate, withhold, predict

As a first application, calibrate a bottom-up potential to a deliberately small dimensionless set, for example

$$\mathcal D_{\mathrm{cal}}= \left\{ \frac{T_c}{\sqrt\sigma}, \frac{p(T_i)}{T_i^4}, \frac{\epsilon(T_i)-3p(T_i)}{T_i^4} \right\}_{i\in I_{\mathrm{train}}},$$

where $\sigma$ is a declared zero-temperature scale and the covariance of the lattice inputs is retained. Fix the energy scale once. Then predict, without refitting, a withheld observable such as the interaction measure at temperatures $I_{\mathrm{test}}$, a screening mass, a susceptibility, or the shape of $\zeta/s(T)$.

For transport, specify the fluctuation action and normalization in addition to the background. The equation of state constrains the equilibrium geometry but does not uniquely fix the Maxwell, flavor, or higher-derivative couplings governing conductivity, diffusion, or viscosity. A recent Bayesian bottom-up study illustrates posterior propagation into several plasma transport quantities Chen et al. 2025; agreement in its calibrated regime remains evidence for that parameterized model and likelihood, not a microscopic derivation of QCD transport.

The comparison should report at least

• the calibrated and withheld data separately;
• dimensionless units and the map to physical units;
• posterior, numerical, derivative-truncation, and model-discrepancy uncertainties;
• the temperature and chemical-potential domain;
• whether the flavor sector is probe or backreacted;
• the large-$N$ and strong-coupling limits.

Two equally good fits as the adversary

Choose two potentials, $V_A(\lambda)$ and $V_B(\lambda)$, that fit the same training equation of state within covariance. Propagate both without refitting to a withheld susceptibility and a transport coefficient. If their predictions disagree beyond numerical error, the training set did not identify the bulk potential or the transport sector. The spread is model uncertainty; selecting the visually smoother potential does not remove it.

The same test applies to a top-down/bottom-up comparison. Similar entropy curves may coexist with different flavor content, anomaly coefficients, or finite-coupling corrections. Agreement across independently sensitive observables can support a common mechanism or phenomenological universality class. It still does not show that either bulk geometry is the unique dual of QCD matter.

Weak-coupling, lattice, effective-theory, hydrodynamic, and experimental analyses provide independent controls in their own regimes. A holographic result is strongest when it states where those regimes overlap or fail, rather than treating strong coupling as a license to ignore them.

Exercises

Why does calibrating $V(\lambda)$ to the equation of state not determine the electric conductivity?

Solution

The equilibrium metric and dilaton constrain thermodynamics, whereas conductivity also depends on the bulk gauge kinetic function, current normalization, and charged matter. Different gauge sectors can live on the same calibrated background and give different current correlators.

The chapter overview contains the structure diagram and validity and failure diagram. They are embedded there once so that their shared chapter-level context is not repeated on every article.

For the chapter-wide comparison of assumptions, counterevidence, falsifiers, and claim ceilings, see the claim-domain table.

References

• Chen, Bing, Liqiang Zhu, Xun Chen, Defu Hou, and Xurong Chen. “Transport Properties of QGP within a Bayesian Holographic QCD Model.” arXiv:2508.16167 [hep-ph] (2025). arXiv.
• Gürsoy, Umut, and Elias Kiritsis. “Exploring Improved Holographic Theories for QCD: Part I.” Journal of High Energy Physics 2008, 032 (2008). DOI.
• Sakai, Tadakatsu, and Shigeki Sugimoto. “Low Energy Hadron Physics in Holographic QCD.” Progress of Theoretical Physics 113, 843–882 (2005). DOI.