Holographic Higher-Derivative Entropy Functionals

When the bulk action contains higher-curvature interactions, the holographic entropy functional is obtained from the response of the action to a replica cone; it is not generally area and is not generally the stationary-black-hole Wald formula. For a generic non-Killing entangling surface, terms quadratic in its extrinsic curvatures accompany the Wald-like term. They vanish on a bifurcation surface or a sufficiently symmetric RT surface, which explains why a pure Wald calculation can pass special checks while failing generically. We use a Euclidean replica saddle continued from a static Lorentzian asymptotically AdS state and treat higher-derivative couplings perturbatively within gravitational EFT.

Required background. Noether-charge and higher-curvature entropy supplies the stationary-horizon functional, and RT supplies the Einstein limit.

Helpful background. Gravity EFT power counting fixes the controlled coupling expansion, and conical entropy explains the replica variation.

The replica functional for curvature actions

Consider a Euclidean action whose local Lagrangian $\mathcal L(g,R_{abcd})$ contains no derivatives of the Riemann tensor. Near a codimension-two surface, choose complex normal directions $z,\bar z$ and tangential indices $i,j$. In Dong’s conical expansion, the entropy takes the structural form

$$\begin{aligned} S_{\rm grav}=2\pi\int_\gamma d^{d-1}y\sqrt h \bigg[ &\frac{\partial\mathcal L}{\partial R_{z\bar z z\bar z}}\\ &+\sum_\alpha \left( \frac{\partial^2\mathcal L} {\partial R_{zizj}\,\partial R_{\bar zk\bar zl}} \right)_\alpha \frac{8K_{z,ij}K_{\bar z,kl}}{q_\alpha+1} \bigg]. \end{aligned}$$

The replica weights $q_\alpha$ are fixed by how the corresponding curvature monomial scales in the regulated cone. The first line is Wald-like; the second is the “anomaly” or extrinsic-curvature contribution. Overall signs are fixed by the Euclidean action and binormal convention, and must reproduce $\operatorname{Area}/4G_N$ for Einstein–Hilbert gravity. Dong derived this formula for general higher-derivative theories, and Camps independently obtained the curvature-squared case (Dong 2014, eqs. (1.3) and (3.29); Camps 2014, §§2–3).

If the action includes derivatives of curvature, additional terms involving normal derivatives and the detailed cone regulator appear. Boundary terms and field redefinitions must be transformed with the action. One cannot insert a preferred Lagrangian representative into a partial formula and expect a field-redefinition-invariant entropy.

A symmetric curvature-squared calculation

Take

$$I=\frac{1}{16\pi G_N}\int d^{d+1}x\sqrt g \left[R-2\Lambda+\alpha L^2R_{abcd}R^{abcd}\right], \qquad |\alpha|\ll1.$$

On constant-curvature Euclidean AdS of radius $L_{\rm eff}$,

$$R_{abcd}=-\frac{1}{L_{\rm eff}^2} (g_{ac}g_{bd}-g_{ad}g_{bc}).$$

Choose the hemispherical surface dual to a vacuum ball. It is the bifurcation surface of an AdS-Rindler Killing horizon, so

$$K_{z,ij}=K_{\bar z,ij}=0.$$

The extrinsic term vanishes. Differentiating the Lagrangian gives

$$\frac{\partial\mathcal L}{\partial R_{abcd}} =\frac{1}{32\pi G_N} (g^{ac}g^{bd}-g^{ad}g^{bc}) +\frac{\alpha L^2}{8\pi G_N}R^{abcd}.$$

Contracting with the normalized binormal yields, to first order in $\alpha$,

$$S_{\rm sym} =\frac{\operatorname{Area}(\gamma)}{4G_N} \left(1-4\alpha\frac{L^2}{L_{\rm eff}^2}\right) +O(\alpha^2),$$

before including the $O(\alpha)$ shift of $L_{\rm eff}$, the regulated area, and the relation between bare and physical couplings. Because the Einstein surface extremizes the zeroth-order area, its $O(\alpha)$ displacement does not contribute to the entropy at first order; evaluating the correction on the unperturbed symmetric surface is sufficient. This is the controlled first application.

For Gauss–Bonnet/Lovelock combinations, the same cone calculation reorganizes into the Jacobson–Myers intrinsic-curvature functional (with its boundary term). On the symmetric surface it agrees with the stationary result; away from symmetry, that agreement follows from the full Lovelock functional, not from discarding extrinsic data by hand (Hung, Myers, and Smolkin 2011, §§4–5).

A direct falsifier for pure Wald entropy

Now perturb the anchor or the state so the extremal surface has nonzero traceless extrinsic curvature. In the $R_{abcd}R^{abcd}$ theory,

$$\frac{\partial^2\mathcal L} {\partial R_{zizj}\,\partial R_{\bar zk\bar zl}} \neq0.$$

Therefore the conical entropy contains a term of the form

$$\Delta S_{\rm extr} \propto \frac{\alpha L^2}{G_N} \int_\gamma\sqrt h\, K_{z,ij}K_{\bar z}^{\,ij},$$

with the exact coefficient and trace combination fixed by the action and replica weights. A pure Wald prescription predicts no such dependence. Vary the shape while holding the ambient curvature at the surface fixed: the Wald term is unchanged at this order, while $\Delta S_{\rm extr}$ changes. The replica effective action therefore directly falsifies pure Wald entropy for the generic non-Killing surface.

The exception is diagnostic, not contradictory. On a Killing bifurcation surface, both extrinsic curvatures vanish and Wald is recovered. On a generic entangling surface they do not, so the full functional must be extremized.

EFT scope and saddle selection

Higher-derivative couplings are controlled only below the EFT cutoff and at an order consistent with all operators of the same dimension. The surface equation follows by varying the complete entropy functional. One then compares every anchored, homologous stationary surface using the same counterterms. Treating the higher-curvature term exactly while omitting equally important EFT operators is not a precision improvement.

Bulk loops add renormalized matter entropy and shift the surface again; that is the distinct FLM/QES problem. Replica derivations explains which cone and continuation assumptions underwrite the functional.

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

• Camps, J. (2014). “Generalized entropy and higher derivative gravity.” Journal of High Energy Physics 2014(3), 070. DOI.
• Dong, X. (2014). “Holographic entanglement entropy for general higher derivative gravity.” Journal of High Energy Physics 2014(1), 044. DOI.
• Hung, L.-Y., Myers, R. C., and Smolkin, M. (2011). “On holographic entanglement entropy and higher curvature gravity.” Journal of High Energy Physics 2011(4), 025. DOI.