In the strict large-N limit, holographic thermal retarded correlators are typigcally meromorphic, with their late-time behavior determined by isolated quasinormal mode (QNM) poles. We show that the first bulk-loop correction, of order 1/N2, qualitatively changes this structure. In a scalar theory on the planar AdS5 black brane, a one-loop bulk diagram generates poles at sums of two QNM frequencies at fixed loop momentum q; integration over the continuous momentum turns these QNM-pair poles into branch cuts. At vanishing external spatial momentum, for pairs built from QNM branches with opposite real parts, the leading light-cone contributions cancel at large momentum, and the residual QNM-pair frequency scales as q−1/3. Consequently, an infinite family of cuts accumulates at zero frequency, while all tree-level QNM poles remain at finite nonzero frequencies. The first finite-N correction therefore closes the gap set by the classical QNM spectrum. Competition between the decreasing QNM-pair damping and thermal Boltzmann suppression produces a moving saddle and a finite-N contribution G(t)∼N−2tαe−Ct3/4,C>0, which decays more slowly than the leading classical QNM contribution. Thus the large-N expansion is nonuniform at late times, and bulk quantum fluctuations generate a distinct late-time regime beyond the classical QNM decay.
Introduction and review of the results
The late-time decay of small perturbations around thermal equilibrium is controlled by the analytic structure of retarded correlation functions. In holography, this structure has been understood predominantly in the strict large-N limit, where the dual bulk description reduces to classical gravity. In that limit the response is organized by the quasinormal modes (QNMs) of the black brane: ingoing normalizable bulk fluctuations appear as poles of boundary retarded correlators, and their imaginary parts set the characteristic damping scales of linear response [1, 2, 3, 4, 5, 6, 7]1. Beyond the strict large-N limit, however, interactions among bulk fluctuations and the associated quantum effects can qualitatively modify this analytic structure. This leads to the questions that motivate this work: what analytic structure supplements the classical QNM poles once such finite-N effects are included, and which part of that structure controls the late-time response?
Recent work has substantially sharpened the large-N picture. Exact and semiclassical thermal correlators and their analytic structure [32, 33, 34, 35, 36], thermal product representations [37, 38], and KMS and spectral-duality constraints [39, 40] have made increasingly explicit how QNM spectra encode the analytic structure of holographic response. More broadly, studies of holographic retarded correlators at complex frequency and momentum have revealed nontrivial structures and constraints on their pole trajectories, including relations to near-horizon bulk dynamics [41, 42]. Related reconstruction methods have shown that, in suitable cases, substantial information about the QNM spectrum and correlator can be recovered from pole-skipping data [43]. Closely related studies have connected singularities of thermal correlators to bulk causal structure, OPE data, and black-hole singularities [44, 45, 46, 47]. Nonanalytic structures can also emerge in special classical limits: in near-extremal backgrounds, nonhydrodynamic poles can condense into a branch cut as T→0[48]. Most of these developments concern the leading large-N, classical-gravity regime. What remains much less understood is the new analytic structure generated by the first bulk loop.
It is useful to distinguish the coupling and 1/N expansions. At leading order in 1/N, finite-coupling effects are encoded by stringy or higher-derivative corrections to the bulk action and can modify the classical QNM spectrum already at tree level [49, 50, 51, 52]. Finite-N effects instead organize the bulk loop expansion [53, 54]. These are independent expansions: a higher-derivative bulk theory can itself receive quantum-loop corrections.
With this distinction in mind, weakly coupled finite-temperature field theory provides a useful point of comparison. There, loops of thermal quasiparticles with continuous relative momentum generate multiparticle branch cuts [55]. At strong coupling, a quasiparticle description need not exist; in the classical holographic regime, thermal response is instead organized by isolated QNM poles, and the contrast between weak-coupling branch cuts and strong-coupling QNM poles has been emphasized in Refs. [56, 57]. The question we address is how robust this pole-based description remains once the first bulk quantum correction is included.
A closely related holographic precedent is the one-loop origin of hydrodynamic long-time tails. Caron-Huot and Saremi showed that bulk loops reproduce the familiar power-law tail associated with hydrodynamic fluctuations [58]; see also Refs. [59, 60, 61, 62, 63]. This demonstrates that bulk quantum fluctuations can generate nonanalytic structure beyond the classical poles. Their analysis, however, concerns gapless hydrodynamic modes and the resulting singular behavior near ω=0. Here we instead ask how a bulk loop built from gapped nonhydrodynamic QNMs modifies the analytic structure of a thermal retarded correlator.
To study this problem, we formulate the one-loop retarded correlator using the real-time Schwinger–Keldysh formalism. Thermal bulk propagators and general real-time gauge/gravity prescriptions were established Refs. [64, 65, 66], followed by holographic Schwinger–Keldysh effective theories and open-system formulations [67, 68, 69]. Thermal higher-point functions were constructed directly from Schwinger–Keldysh contours in Ref. [70]. More recently, an exterior effective field theory description of real time black hole dynamics was developed [71] and extended to bulk loop diagrams [72]. These developments provide the framework for computing real-time bulk quantum corrections. We use this framework below to determine the analytic structure generated by the QNM spectrum at first loop order.
A direct calculation of this problem is considerably harder than the classical QNM problem. Let us suppose that the bulk theory contains interactions capable of generating the one-loop correction shown schematically in Figure 1. Two external retarded legs connect the boundary insertions to bulk interaction points at radial positions z and z′, while two real-time bulk propagators run around the loop. The calculation therefore involves integrations over the loop frequency and momentum, together with independent integrations over the two bulk radial positions. The quantity of interest is the resulting boundary correlator.
Figure 1. Schematic bulk one-loop contribution to the boundary retarded correlator. The external retarded bulk-to-boundary propagators KR(z) and KR(z′) connect the boundary insertions to bulk vertices at radial positions z and z′, while ΣR(z,z′) denotes the retarded bulk self-energy kernel generated by the loop. The two radial positions are integrated independently and are drawn at similar depths only for visual clarity.
The singular part of the correlator can be isolated more efficiently. At fixed loop momentum, the two internal propagators have QNM poles in the complex loop-frequency plane. As the external frequency is analytically continued, a pole from each propagator can approach the loop-frequency contour from opposite sides. When the two poles meet, the contour becomes pinched and the correlator develops a nonanalyticity. This is the thermal analogue of the familiar Landau-pinch mechanism in finite-temperature field theory [55, 73, 74]. The pinch produces a pole at the sum of the two QNM frequencies. The remaining momentum integration determines the resulting branch-cut structure, while the radial integrations determine whether the singular contribution survives in the boundary correlator and with what strength.
For simplicity, we first specialize to vanishing external spatial momentum, κ=0.2 The two internal propagators then carry momenta q and −q, with q=∣q∣. We denote their QNM frequencies by ωn±(q), where n=0,1,… labels the overtone and the signs distinguish the positive and negative real frequency branches. The possible pinches are organized by the pair frequencies
Ωmnστ(q)=ωmσ(q)+ωnτ(q),σ,τ=±.(1)
As the loop momentum varies, these pair frequencies trace curves in the complex-frequency plane. These curves identify candidate branch cuts, but not every point on them is a branch point of the boundary correlator. After the momentum integration, branch points occur at special values of q, such as endpoints of the integration domain or stationary points of Ωmnστ(q), provided the corresponding boundary coefficient is nonzero. Varying the signs and overtone indices organizes the resulting pair-frequency loci into the QNM-pair networkc shown in Figure 2.
Figure 2. Representative QNM-pair branch cuts at vanishing external spatial momentum, κ=0. Blue crosses denote the classical QNM frequencies ωn±(0), while black dots mark the q=0 endpoints of the branch cuts. The solid zigzags show the numerically determined cuts for 0≤q≤100, parameterized by the loop momentum q; faint dotted zigzags show their large-q asymptotic continuation. The q=0 endpoints are square-root branch points of the boundary correlator. The diagonal mixed-sign cuts lie on the negative imaginary axis, with the fundamental 00+− channel emphasized; off-diagonal mixed-sign cuts occur in reflected pairs and accumulate at the origin as q→∞. The red dashed lines bound the large-q asymptotic sector, ∣θ∣<30∘, derived in Eq. (5). The top-left inset resolves the closely spaced same-sign branch points.
The two sign sectors have qualitatively different geometries. Same-sign pairs generate the outer families shown in the figure. For mixed-sign pairs, the reflection relation ωn−(q)=−[ωn+(q)]∗ gives Ωmn+−(q)=ωm+(q)−[ωn+(q)]∗. The real parts therefore cancel for m=n, placing the diagonal mixed channels on the negative imaginary axis. For m=n, this cancellation is incomplete, and exchanging the two indices reflects the pair-frequency curve across that axis. The central network thus contains both overlapping diagonal channels and reflected off-diagonal families.
ω≡πTωphys,κ≡πTk,q≡πTqphys,τ≡πTt,(2)
where κ denotes the external momentum and q the loop momentum. The restriction to κ=0 is made to expose the structure in its simplest form. We comment on κ=0 in § 6.
A further organizing principle emerges at large loop momentum. For each fixed overtone, the large-q QNM asymptotics derived by Festuccia and Liu [7] take the form
ωn+(q)=q+2λne−iπ/3q−1/3+O(q−5/3),(3)
where 0<λ0<λ1<⋯ are the positive eigenvalues defined in App. § 12. Using the reflected negative real frequency branch then gives, for a mixed-sign pair,
Consequently, every mixed-sign pair approaches ω=0 as q→∞, so the origin is an accumulation point of the entire mixed-sign tower. The direction of approach is also constrained. If θmn is measured from the negative imaginary axis,
tanθmn=3(λm+λn)λm−λn,∣θmn∣<6π.(5)
Thus all fixed-index mixed-sign channels approach the origin within the 60° sector bounded by the red dashed rays in Figure 2.
The accumulation at zero frequency does not imply that all mixed-sign channels contribute equally to late-time relaxation. For every mixed-sign channel, approaching the origin requires increasing the loop momentum: this reduces the QNM-pair damping, but at the same time suppresses the thermal weight through the Boltzmann factor e−βq=e−πq. The late-time response is therefore governed by a competition between these two effects. Their balance selects a momentum scale that grows as τ3/4 and produces a stretched-exponential decay, slower than any ordinary exponential e−γτ with fixed γ>0. The quantitative suppression depends on the QNM pair, and the saddle-point analysis gives
Gmnmin(τ)∼ϵAmnτν/2+1/8e−Cmnτ3/4.(6)
Here Amn determines the strength of the contribution from the (m,n) mixed-sign channel and is obtained from the QNM residues and radial integrations. This asymptotic form applies whenever Amn=0. In Eq. (6), ReCmn is positive and controls the stretched-exponential suppression of each channel. We show below that the fundamental 00+− channel has the smallest ReCmn, and therefore decays most slowly at asymptotically late times. Its special role comes from this ordering of decay rates, not from being the only cut that reaches zero frequency.
To summarize, at leading order in the large-N expansion, the scalar correlator is meromorphic and all of its isolated QNM poles remain a finite distance from zero frequency. At the first finite-N correction, branch cuts appear and extend to zero frequency. Thus the analytic structure changes qualitatively already at order 1/N2, closing the gap set by the classical QNM spectrum. As a result, there is a late-time regime in which the stretched-exponential decay generated by the one-loop branch-cut contribution dominates over the exponential decay governed by the leading classical QNM contribution.
The rest of the paper is organized as follows. § 2 introduces the holographic model and the one-loop retarded correlator. In § 3 we derive the QNM-pair singularity condition and the corresponding boundary coefficient. We than analyze the lowest mixed-sign channel and the resulting branch-cut network. In § 4 we study the large-momentum accumulation of the mixed-sign cuts near ω=0, and in § 5 we determine the resulting late-time response. § 6 discusses the implications and possible extensions. Technical details and numerical checks are collected in the appendices.
Holographic setup and one-loop retarded correlator
We now formulate the one-loop problem in real-time holography. We first specify the bulk model and the retarded and thermal propagators entering the loop, and then project the first bulk self-energy insertion onto the boundary correlator. This construction also makes explicit how the QNM poles of the internal propagators arise from the bulk radial problem and enter the loop dynamics.
Bulk model and linear propagators
We work on the planar AdS5–Schwarzschild black brane, whose horizon radius is related to the temperature by zh=πT1. We introduce dimensionless horizon coordinates
z=zhzphys,τ=zhtphys=πTtphys,x=zhxphys.
In these coordinates the horizon is at z=1, and zh has been scaled out of the metric. Setting the AdS radius to one, we obtain
ds2=z21[−f(z)dτ2+dx2+f(z)dz2],f(z)=1−z4.(7)
The AdS boundary is at z=0. Since the physical Euclidean time has period 1/T=πzh, the dimensionless Euclidean coordinate τE has period π.
On this background we consider a scalar field ϕ with a cubic self-interaction,
S=−A∫d5x−g[21(∇ϕ)2+21m2ϕ2+3!gϕ3+⋯].(8)
We choose m2=1 as a representative positive scalar mass; this keeps the spectrum gapped and avoids special zero-mode or boundary-degeneracy effects. The ellipsis denotes additional local interactions and counterterms needed to complete and renormalize the model. In particular, we impose the tadpole renormalization condition such that the scalar one-point function vanishes and ϕ=0 remains the background about which the two-point function is computed. These local terms do not affect the nonlocal singular structure studied below.
References
[1]G.T. Horowitz and V.E. Hubeny, Quasinormal modes of AdS black holes and the approach to thermal equilibrium, Phys. Rev. D 62 (2000) 024027 [hep-th/9909056].
[2]D. Birmingham, I. Sachs and S.N. Solodukhin, Conformal field theory interpretation of black hole quasinormal modes, Phys. Rev. Lett. 88 (2002) 151301 [hep-th/0112055].pubmed.ncbi.nlm.nih.gov/11955184
[3]D.T. Son and A.O. Starinets, Minkowski-space correlators in AdS/CFT correspondence: recipe and applications, JHEP 09 (2002) 042 [hep-th/0205051].
[4]A. Núñez and A.O. Starinets, AdS/CFT correspondence, quasinormal modes, and thermal correlators in N = 4 SYM, Phys. Rev. D 67 (2003) 124013 [hep-th/s].DOI
[5]P.K. Kovtun and A.O. Starinets, Quasinormal modes and holography, Phys. Rev. D 72 (2005) 086009 [hep-th/0506184].DOI
[6]E. Berti, V. Cardoso and A.O. Starinets, Quasinormal modes of black holes and black branes, Class. Quant. Grav. 26 (2009) 163001 [arXiv:0905.2975].DOI
[7]G. Festuccia and H. Liu, A Bohr–Sommerfeld quantization formula for quasinormal frequencies of AdS black holes, Adv. Sci. Lett. 2 (2009) 221 [arXiv:0811.1033].DOI
[8]M. Kaminski, K. Landsteiner, J. Mas, J.P. Shock and J. Tarrio, Holographic Operator Mixing and Quasinormal Modes on the Brane, JHEP 02 (2010) 021 [arXiv:0911.3610].DOI
[9]M. Ammon, M. Kaminski, R. Koirala, J. Leiber and J. Wu, Quasinormal modes of charged magnetic black branes & chiral magnetic transport, JHEP 04 (2017) 067 [arXiv:1701.05565].DOI
[10]M. Baggioli, S. Grieninger and L. Li, “Magnetophonons & type-B Goldstones from hydrodynamics to holography,” JHEP 09 (2020) 037 [arXiv:2005.01725 [hep-th]].DOI
[11]R.A. Davison, S.A. Gentle and B. Gouteraux, Slow Relaxation and Diffusion in Holographic Quantum Critical Phases, Phys. Rev. Lett. 123 (2019) 141601 [arXiv:1808.05659].DOI
[12]B. Meiring, I. Shyovitz, S. Waeber and A. Yarom, “Multiply charged magnetic black branes,” JHEP 06 (2024), 196 [arXiv:2312.02802 [hep-th]].arxiv.org/abs/2312.02802
[13]X.-J. Wang and W.-J. Li, “Holographic phonons by gauge-axion coupling,” JHEP 07 (2021) 131 [arXiv:2105.07225 [hep-th]].DOI
[14]X.-H. Ge, J.-R. Sun, Y. Tian, X.-N. Wu and Y.-L. Zhang, “Holographic interpretation of acoustic black holes,” Phys. Rev. D 92 (2015) 084052 [arXiv:1508.01735 [hep-th]].DOI
[15]L. A. H. Mamani, D. Hou and N. R. F. Braga, “Melting of heavy vector mesons and quasinormal modes in a finite density plasma from holography,” Phys. Rev. D 105 (2022) 126020 [arXiv:2204.08068 [hep-ph]].DOI
[16]R. Chen, D. Li and M. Huang, “Thermal properties of the scalar glueballs from holography,” Phys. Rev. D 112 (2025) 094021 [arXiv:2507.20309 [hep-ph]].arxiv.org/abs/2507.20309
[17]H.-B. Li, T.-T. Hu, B.-S. Song, S. Sun and Y.-Q. Wang, “Deforming black holes with even multipolar differential rotation boundary,” JHEP 06 (2019) 126 [arXiv:1903.11967 [gr-qc]].DOI
[18]S. He and Y. Li, “Holographic Euclidean thermal correlator,” JHEP 03 (2024) 024 [arXiv:2308.13518 [hep-th]].arxiv.org/abs/2308.13518
[19]Q. Tan, W.-D. Guo, Y.-P. Zhang and Y.-X. Liu, “Characteristic modes of a thick brane: Resonances and quasinormal modes,” Phys. Rev. D 109 (2024) 024017 [arXiv:2304.09363 [gr-qc]].DOI
[20]S. I. Finazzo, R. Rougemont, M. Zaniboni, R. Critelli and J. Noronha, “Critical behavior of non-hydrodynamic quasinormal modes in a strongly coupled plasma,” JHEP 01 (2017) 137 [arXiv:1610.01519 [hep-th]].
[21]U. Gürsoy, A. Jansen and W. van der Schee, “A new dynamical instability in asymptotically Anti-de-Sitter spacetime,” Phys. Rev. D 94 (2016) 061901 [arXiv:1603.07724 [hep-th]].DOI
[22]D. Arean, D. Garcia-Farinas and K. Landsteiner, Pseudospectra of Holographic Quasinormal Modes, JHEP 12 (2023) 187 [arXiv:2307.08751].DOI
[23]M.P. Heller, A. Serantes, M. Spalinski, V. Svensson and B. Withers, Convergence of hydrodynamic modes: insights from kinetic theory and holography, SciPost Phys. 10 (2021) 123 [arXiv:2012.15393].DOI
[25]S. Grozdanov, P.K. Kovtun, A.O. Starinets and P. Tadić, Convergence of the Gradient Expansion in Hydrodynamics, Phys. Rev. Lett. 122 (2019) 251601 [arXiv:1904.01018].DOI
[26]S. Grozdanov, P.K. Kovtun, A.O. Starinets and P. Tadić, The complex life of hydrodynamic modes, JHEP 11 (2019) 097 [arXiv:1904.12862].DOI
[27]N. Abbasi and S. Tahery, Complexified quasinormal modes and the pole-skipping in a holographic system at finite chemical potential, JHEP 10 (2020) 076 [arXiv:2007.10024].DOI
[28]R. Emparan and K. Tanabe, “Universal quasinormal modes of large D black holes,” Phys. Rev. D 89 (2014) no.6, 064028 [arXiv:1401.1957 [hep-th]].DOI
[29]M.P. Heller, Holography, Hydrodynamization and Heavy-Ion Collisions, Acta Phys. Polon. B 47 (2016) 2581 [arXiv:1610.02023].DOI
[30]W. Florkowski, M.P. Heller and M. Spalinski, New theories of relativistic hydrodynamics in the LHC era, Rept. Prog. Phys. 81 (2018) 046001 [arXiv:1707.02282].DOI
[31]H. L. Jia, W. D. Guo, Y. T. Gu and Y. X. Liu, “Real part emergence in purely imaginary quasinormal modes in perturbed de Sitter braneworlds,” JHEP 10 (2026), 265 [arXiv:2606.04424 [gr-qc]].arxiv.org/abs/2606.04424
[32]A. Bhatta and T. Mandal, Exact thermal correlators of holographic CFTs, JHEP 02 (2023) 222 [arXiv:2211.02449].DOI
[33]H.F. Jia and M. Rangamani, Holographic thermal correlators and quasinormal modes from semiclassical Virasoro blocks, JHEP 12 (2024) 047 [arXiv:2408.05208].arxiv.org/abs/2408.05208
[34]I. Burić, I. Gusev and A. Parnachev, Thermal holographic correlators and KMS condition, arXiv:2505.10277.arxiv.org/abs/2505.10277
[35]P. Arnaudo and B. Withers, Analytic structure of holographic thermal correlators from Fourier series, JHEP 06 (2026) 205 [arXiv:2603.13469].arxiv.org/abs/2603.13469
[36]H.F. Jia and M. Rangamani, Exact holographic thermal spectral functions: OPE, non-perturbative corrections, and black hole singularity, arXiv:2604.10803.arxiv.org/abs/2604.10803
[37]M. Dodelson, C. Iossa, R. Karlsson and A. Zhiboedov, A thermal product formula, JHEP 01 (2024) 036 [arXiv:2304.12339].arxiv.org/abs/2304.12339
[38]J. Bhattacharya, N. Padhi, A. Sharma and S. Singha, Thermal product formula for shear modes, JHEP 08 (2025) 170.DOI
[39]S. Grozdanov and M. Vrbica, Duality constraints on thermal spectra of 3d conformal field theories and 4d quasinormal modes, Phys. Rev. Lett. 133 (2024) 211601 [arXiv:2406.19790].DOI
[40]S. Grozdanov and M. Vrbica, Thermal field theory correlators in the large-N limit and the spectral duality relation, JHEP 02 (2026) 106.arxiv.org/abs/2509.18074
[41]M. Blake, R. A. Davison, S. Grozdanov and H. Liu, “Many-body chaos and energy dynamics in holography,” JHEP 10 (2018), 035, [arXiv:1809.01169 [hep-th]].DOI
[42]M. Blake, R. A. Davison and D. Vegh, “Horizon constraints on holographic Green’s functions,” JHEP 01 (2020), 077, [arXiv:1904.12883 [hep-th]].DOI
[43]S. Grozdanov, T. Lemut and J. F. Pedraza, “Reconstruction of the quasinormal spectrum from pole skipping,” Phys. Rev. D 108 (2023) no.10, L101901 [arXiv:2308.01371 [hep-th]].DOI
[44]M. Dodelson, C. Iossa, R. Karlsson, A. Lupsasca and A. Zhiboedov, Black hole bulk-cone singularities, JHEP 07 (2024) 046.DOI
[45]N. Čeplak, H. Liu, A. Parnachev and S. Valach, Black hole singularity from OPE, JHEP 10 (2024) 105 [arXiv:2404.17286].arxiv.org/abs/2404.17286
[46]N. Afkhami-Jeddi, S. Caron-Huot, J. Chakravarty and A. Maloney, Imprint of the black hole singularity on thermal two-point functions, Phys. Rev. D 114 (2026) 025005 [arXiv:2510.21673].
[47]S. Grozdanov, S. Valach and M. Vrbica, Bouncing geodesics, black hole singularities, and singularities of thermal correlators, JHEP 08 (2026) 022.
[48]B. Goutéraux, D. M. Ramirez, M. Sanchez-Garitaonandia and C. Supiot, “Near-extremal holographic charge correlators,” Phys. Rev. D 114 (2026) no.8, 086006 [arXiv:2506.11974 [hep-th]].arxiv.org/abs/2506.11974
[49]S.A. Stricker, Holographic thermalization in N = 4 super Yang–Mills theory at finite coupling, Eur. Phys. J. C 74 (2014) 2727 [arXiv:1307.2736].DOI
[50]S. Waeber, A. Schäfer, A. Vuorinen and L.G. Yaffe, Finite coupling corrections to holographic predictions for hot QCD, JHEP 11 (2015) 087 [arXiv:1509.02983].DOI
[51]S. Grozdanov, N. Kaplis and A.O. Starinets, From strong to weak coupling in holographic models of thermalization, JHEP 07 (2016) 151 [arXiv:1605.02173].DOI
[52]A. Buchel, “Sensitivity of holographic N=4 SYM plasma hydrodynamics Phys. Rev. D 98 (2018) no.6, 061901 [arXiv:1807.05457 [hep-th]].DOI
[53]O. Aharony, S.S. Gubser, J.M. Maldacena, H. Ooguri and Y. Oz, Large N field theories, string theory and gravity, Phys. Rept. 323 (2000) 183–386 [hep-th/9905111].DOI
[54]O. Aharony, L.F. Alday, A. Bissi and E. Perlmutter, Loops in AdS from conformal field theory, JHEP 07 (2017) 036 [arXiv:1612.03891].DOI
[55]H.A. Weldon, Branch cuts due to finite-temperature quasiparticles, Phys. Rev. D 58 (1998) 105002 [hep-ph/9806325].DOI
[56]S. A. Hartnoll and S. P. Kumar, “AdS black holes and thermal Yang–Mills correlators,” JHEP 12 (2005) 036 [arXiv:hep-th/0508092].DOI
[57]D. F. Hou, J. R. Li, H. Liu and H. C. Ren, “The Momentum analyticity of two-point correlators from perturbation theory and AdS/CFT,” JHEP 07 (2010), 042 [arXiv:1003.5462 [hep-ph]].DOI
[58]S. Caron-Huot and O. Saremi, Hydrodynamic long-time tails from anti-de Sitter space, JHEP 11 (2010) 013 [arXiv:0909.4525].
[59]P. Kovtun and L.G. Yaffe, Hydrodynamic fluctuations, long-time tails, and supersymmetry, Phys. Rev. D 68 (2003) 025007 [hep-th/0303010].DOI
[60]P. Kovtun, G.D. Moore and P. Romatschke, The stickiness of sound: an absolute lower limit on viscosity and the breakdown of second-order relativistic hydrodynamics, Phys. Rev. D 84 (2011) 025006.DOI
[61]P. Kovtun, Lectures on hydrodynamic fluctuations in relativistic theories, J. Phys. A 45 (2012) 473001 [arXiv:1205.5040].DOI
[62]X. An, G. Başar, M. Stephanov and H.-U. Yee, Evolution of non-Gaussian hydrodynamic fluctuations, Phys. Rev. Lett. 127 (2021) 072301 [arXiv:2009.10742].DOI
[64]C.P. Herzog and D.T. Son, Schwinger–Keldysh propagators from AdS/CFT correspondence, JHEP 03 (2003) 046 [hep-th/2].DOI
[65]K. Skenderis and B.C. van Rees, Real-time gauge/gravity duality, Phys. Rev. Lett. 101 (2008) 081601 [arXiv:0805.0150].DOI
[66]K. Skenderis and B.C. van Rees, Real-time gauge/gravity duality: prescription, renormalization and examples, JHEP 05 (2009) 085 [arXiv:0812.2909].DOI
[67]J. de Boer, M.P. Heller and N. Pinzani-Fokeeva, Holographic Schwinger–Keldysh effective field theories, JHEP 05 (2019) 188 [arXiv:1812.06093].DOI
[68]C. Jana, R. Loganayagam and M. Rangamani, Open quantum systems and Schwinger–Keldysh holograms, JHEP 07 (2020) 242 [arXiv:2004.02888].DOI
[69]R. Loganayagam, M. Rangamani and J. V i r r u e t a, Holographic open quantum systems: toy models and analytic properties of thermal correlators, JHEP 03 (2023) 153 [arXiv:2211.07683].DOI
[70]C. Pantelidou and B. Withers, Thermal three-point functions from holographic Schwinger–Keldysh contours, JHEP 04 (2023) 050 [arXiv:2211.09140].DOI
[71]R. Loganayagam and G. Martin, An exterior EFT for Hawking radiation, JHEP 06 (2025) 184 [arXiv:2403.10654].arxiv.org/abs/2403.10654
[72]R. Loganayagam, G. Martin and S.K. Sharma, Loops outside a black hole, JHEP 06 (2026) 087.arxiv.org/abs/2509.03656
[73]L.D. Landau, On analytic properties of vertex parts in quantum field theory, Nucl. Phys. 13 (1959) 181.DOI
[74]H.A. Weldon, Simple rules for discontinuities in finite-temperature field theory, Phys. Rev. D 28 (1983) 2007.DOI
[75]J.B. Hartle and S.W. Hawking, Path-integral derivation of black-hole radiance, Phys. Rev. D 13 (1976) 2188.
[76]D. T. Son and D. Teaney, “Thermal Noise and Stochastic Strings in AdS/CFT,” JHEP 07 (2009), 021 [arXiv:0901.2338 [hep-th]].DOI
[77]S.S. Gubser, I.R. Klebanov and A.M. Polyakov, Gauge theory correlators from non-critical string theory, Phys. Lett. B 428 (1998) 105–114 [hep-th/9802109].DOI
[78]E. Witten, Anti-de Sitter space and holography, Adv. Theor. Math. Phys. 2 (1998) 253–291 [hep-th/9802150].arxiv.org/abs/hep-th/9802150
[79]L. Susskind and E. Witten, “The Holographic bound in anti-de Sitter space,” [arXiv:hep-th/9805114 [hep-th]].DOI
[80]T. Banks, M. R. Douglas, G. T. Horowitz and E. J. Martinec, “AdS dynamics from conformal field theory,” [arXiv:hep-th/9808016 [hep-th]].DOI
[81]D. Harlow and D. Stanford, “Operator Dictionaries and Wave Functions in AdS/CFT and dS/CFT,” [arXiv:1104.2621 [hep-th]].
[82]A. Kamenev and A. Levchenko, “Keldysh technique and nonlinear sigma-model: Basic principles and applications,” Adv. Phys. 58 (2009), 197 [arXiv:0901.3586 [cond-mat.other]].
[83]M. Le Bellac, Thermal Field Theory, Cambridge University Press, Cambridge (1996).
[84]X. Chen-Lin, L. V. Delacrétaz and S. A. Hartnoll, “Theory of Diffusive Fluctuations,” Phys. Rev. Lett. 122 (2019) no.9, 091602, [arXiv:1811.12540 [hep-th]].
[85]N. Abbasi, M. Kaminski and O. Tavakol, “Theory of Nonlinear Diffusion with a Physical Gapped Mode,” Phys. Rev. Lett. 132 (2024) no.13, 131602 [arXiv:2212.11499 [hep-th]].arxiv.org/abs/2212.11499
[86]R.J. Eden, P.V. Landshoff, D.I. Olive and J.C. Polkinghorne, The Analytic S-Matrix, Cambridge University Press, Cambridge (1966).DOI
[87]H. S. Hannesdottir, A. J. McLeod, M. D. Schwartz and C. Vergu, “Constraints on sequential discontinuities from the geometry of on-shell spaces,” JHEP 07 (2023), 236 [arXiv:2211.07633 [hep-th]].DOI
[88]N. Abbasi, A. Davody and S. Tahery, Correlation functions in stable first-order relativistic hydrodynamics, Phys. Rev. D 109 (2024) 036006 [arXiv:2212.14619].arxiv.org/abs/2212.14619
[89]N. Abbasi, M. Kaminski and D. H. Rischke, “Comparison between causal and acausal diffusion: a Schwinger-Keldysh effective field theory perspective,” JHEP 03 (2026), 156 [arXiv:2506.20500 [hep-th]].arxiv.org/abs/2506.20500
[90]D. L. Wang and S. Pu, “Late-Time Relaxation from Landau Singularities,” [arXiv:2605.04020 [hep-th]].arxiv.org/abs/2605.04020
[92]J. R. V. Crump and J. E. Santos, “After the Fluid: Subexponential Decay in AdS₄,” [arXiv:2605.18978 [hep-th]].arxiv.org/abs/2605.18978
[93]E. Witten, Why does quantum field theory in curved spacetime make sense? And what happens to the algebra of observables in the thermodynamic limit?, arXiv:2112.11614.arxiv.org/abs/2112.11614
[94]J.-M. Schlenker and E. Witten, No ensemble averaging below the black hole threshold, JHEP 07 (2022) 143 [arXiv:2202.01372].DOI
[96]J. Casalderrey-Solana, H. Liu, D. Mateos, K. Rajagopal and U.A. Wiedemann, Gauge/String Duality, Hot QCD and Heavy Ion Collisions, Cambridge University Press (2014) [arXiv:1101.0618].
Arthur Bouffandeau, David Poisbleau, Mohand Ou Idir Ouidja, Patrick Karasinski, Dulce Papy, Charles-Henri Flouzat-Lachaniette, Romain Bosc, Barbara Hersant, Jean-Paul Meningaud, Guillaume Haiat ·