Transition from weak to strong coupling in thermal gauge theories: Lessons from ${\cal N}=4$ SYM Theory

Berndt M\"uller

REVIEW 4 major objections 3 minor 16 references

Transition from weak to strong coupling in thermal gauge theories: Lessons from ${\cal N}=4$ SYM Theory

T0 review · 4 major / 3 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read In thermal N=4 SYM, the weak-to-strong-coupling transition for entropy and three transport coefficients is a broad crossover in the range $3 \lesssim \lambda \lesssim 14$, so at $\lambda \approx 12$ neither expansion is reliable.

desk verdict Useful crossover estimates for N=4 SYM transport coefficients, but the η/s entry rests on an assumed scaling relation, not a direct NLO calculation; the qualitative message survives. read the letter →

arxiv 2507.06845 v2 pith:EELV4OFG submitted 2025-07-09 hep-th hep-phnucl-th

Berndt Müller This is my paper

classification hep-thhep-phnucl-th

keywords N=4supersymmetricYang-Millstheory'tHooftcouplingweak-to-strongtransitionPadéapproximantshearviscosityjetquenchingparameterheavy-quarkdiffusionquark-gluonplasma

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper asks where the weak-coupling and strong-coupling descriptions of a hot gauge plasma lose their validity, using $\mathcal{N}=4$ supersymmetric Yang-Mills (SYM) theory as a controlled test bed. For the normalized entropy density, the kinematic shear viscosity, the heavy-quark diffusion constant, and the transverse momentum broadening parameter $\hat{q}/T^3$, next-to-leading-order results are known on both sides of the coupling axis, and the paper bridges them with Padé approximants. The result is that there is no sharp phase transition as a function of 't Hooft coupling $\lambda$; instead there is a gradual crossover, and all four quantities put that crossover in the region $3 \lesssim \lambda \lesssim 14$. Since $\lambda \approx 12$ is a value often used in applying SYM results to quark-gluon plasma transport, the paper concludes that in this physically relevant window neither the strong-coupling nor the weak-coupling expansion is quantitatively reliable.

What carries the argument

The key instrument is the generalized Padé approximant: a rational function in powers of $\lambda^{1/2}$ whose coefficients are fixed by matching the known next-to-leading-order weak- and strong-coupling expansions. The transition point is then defined by an extremum of the logarithmic slope (for the entropy density) or of the logarithmic curvature (for the three transport coefficients), and the transition width by the points where that derivative has fallen to half its peak value. For the shear viscosity, whose NLO weak-coupling result is not known, the approximant is built from the scaling relation $(\eta/s)(\hat{q}/T^3) \approx \text{const}$, using the NLO $\hat{q}$ expression in its place. On the strong-coupling side the machinery includes the finite-coupling stringy corrections of order $\lambda^{-3/2}$ (and $\lambda^{-1/2}$ for $\hat{q}$) coming from the AdS/CFT gravity dual.

What would settle it

Directly compute the NLO weak-coupling $\eta/s$ in $\mathcal{N}=4$ SYM and test whether $(\eta/s)(\hat{q}/T^3)$ is truly constant in the weak-coupling region; a variation larger than the paper's assumed value would change the $\eta/s$ transition curve and the combined window. Alternatively, a lattice simulation of thermal $\mathcal{N}=4$ SYM at $\lambda$ between 3 and 14 could measure the logarithmic curvature of $\hat{q}/T^3$ or $\eta/s$ and check whether the predicted crossover shape is real.

Watch

Extended reading notes

Core claim

The central discovery is that the border between weak and strong coupling in thermal $\mathcal{N}=4$ SYM is not a point but a broad interval, and that the interval is essentially the same for four quite different quantities. Using the first logarithmic derivative of the interpolated entropy density and the second logarithmic derivative of the interpolated transport coefficients to locate a central pseudocritical coupling $\lambda_c$, the paper obtains $\lambda_c = 3.14$ for $s/s_0$, $5.28$ for $\eta/s$, $11.90$ for $2\pi T D_s$, and $4.36$ for $\hat{q}/T^3$, with transition regions whose half-widths are factors of 2 to 3. Averaged over the four quantities, the full transition region is $3 \lesssim \lambda \lesssim 14$. Because $\lambda \approx 12$ is the value often invoked when transferring SYM results to quark-gluon plasma transport, the paper concludes that this range falls in the crossover and that neither the strong-coupling nor the weak-coupling expansion can be trusted there.

Load-bearing premise

The shear-viscosity interpolation assumes that the combination $(\eta/s)(\hat{q}/T^3)$ stays nearly independent of coupling in weakly coupled $\mathcal{N}=4$ SYM, since the full NLO weak-coupling $\eta/s$ has not been computed; if that scaling relation fails, the $\eta/s$ transition point and the averaged $3 \lesssim \lambda \lesssim 14$ window shift.

Editorial extensions

If this is right

At $\lambda \approx 12$, the value commonly used to model quark-gluon plasma with $\mathcal{N}=4$ SYM, the system sits in the middle of the crossover, so extrapolating either the strong-coupling or the weak-coupling limit to that point is unreliable.
There is no phase transition in 't Hooft coupling: the crossover has a width of a factor of 2 to 3 around $\lambda_c$, so any single 'transition coupling' is a bookkeeping device rather than a sharp threshold.
Averaged over the entropy density and the three transport coefficients, the weak-to-strong transition window is $3 \lesssim \lambda \lesssim 14$.
When the same logic is applied to QCD at $\alpha_s(2\pi T) \approx 0.3$, corresponding to $\lambda \approx 11$, the effective coupling also falls in the crossover, making both perturbative thermal QCD and strong-coupling SYM modeling questionable in that temperature range.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The same Padé-crossing method could be applied to other $\mathcal{N}=4$ SYM quantities, such as bulk viscosity or momentum diffusion in a flowing medium, to test whether the crossover window is universal across observables.
A reader should expect later higher-order calculations to shift $\lambda_c$ within the quoted half-widths, but the broad window between roughly 3 and 14 is likely to survive.
A lattice simulation of thermal $\mathcal{N}=4$ SYM at moderate $\lambda$ could measure the curvature of transport coefficients directly and test whether the Padé-interpolated crossover shape is the real behavior.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Here's the quick take. The paper does one useful thing: it takes the Padé interpolation idea from Andersen et al. [7] and applies it to three transport coefficients in N=4 SYM, giving first estimates of where the weak-to-strong coupling crossover sits for η/s, D, and qhat. The punchline—that at λ≈11-12, the coupling used in many QGP applications, neither the strong nor the weak coupling expansion is reliable—is plausible and worth taking seriously. The author is transparent about the method's arbitrariness: he states that the choice of Padé form and the derivative criterion are not unique, and he reports the sensitivity to one alternative (intersection vs curvature) explicitly.

The main soft spot is the η/s leg. The NLO weak-coupling η/s in N=4 SYM is unknown, and the paper substitutes the scaling relation (η/s)(qhat/T^3) ≈ const, using the known NLO qhat. That is an assumption, not a calculation, and the author says so in the appendix. The stress-test note is right that this makes the η/s λc=5.28 and its width conditional. But the paper already flags this: it points out the intersection criterion would give λc=3.01 for η/s and attributes the discrepancy partly to the substitution. So the reader is not misled.

Is the 3≤λ≤14 headline robust? Mostly. The three other quantities—entropy, qhat, and to a lesser extent D—have λc values in 3–12, and their transition ranges extend to λ≈13 or beyond. The diffusion constant's λc is 11.9, which itself sits right in the interesting region. So even if the η/s entry were discarded, the qualitative conclusion that λ≈12 is inside the crossover rather than deep in either regime would still stand. What is weaker is the "averaging" step: the individual transition ranges are quite different (e.g., D's upper bound is 27.8), and the stated 3–14 is a rough envelope rather than a rigorous derivation. Readers should treat it as an indicative range, not a tight constraint.

The citation pattern is fine; the paper builds on published NLO results and gives credit where due. It is a modest extension of an existing method, not a new framework. But it is a useful numbers paper, and the physics message is timely for the heavy-ion community.

I'd send it to review. A good referee will check the η/s substitution and ask the author to either soften the claim or find a direct NLO result, but the paper is honest, clear, and short. Bring it to reading group if anyone cares about AdS/CFT applications to QGP.

Referee Report

4 major / 3 minor

Summary. This paper studies the weak-to-strong coupling transition in N=4 supersymmetric Yang-Mills theory as a function of the 't Hooft coupling λ. For four quantities with known next-to-leading-order expansions in both regimes—normalized entropy density, shear viscosity η/s, heavy-quark diffusion constant 2πT D_s, and jet quenching parameter qhat/T^3—the author constructs generalized Padé interpolants and locates each transition point λ_c as an extremum of a logarithmic derivative or curvature. The reported λ_c values lie between 3.1 and 11.9, and combining the four transition regions yields an overall estimate 3 ≲ λ ≲ 14. The paper concludes that neither the weak-coupling nor the strong-coupling expansion is reliable in the range around λ ≈ 12 that is often used for quark-gluon plasma applications.

Significance. If the interpolations are trustworthy, the paper provides a compact, physically useful map of where asymptotic expansions fail in a theory that serves as a model for the quark-gluon plasma. The approach is transparent: explicit Padé forms are given, and the inputs are published NLO calculations, which is a strength. The central quantitative claim, however, rests on one input that is not a direct NLO result: the weak-coupling η/s is obtained from an assumed scaling relation (η/s)(qhat/T^3) ≈ const, rather than from a calculated NLO expression for η/s. Because the paper itself demonstrates a large definition-dependence for this quantity (λ_c = 3.01 vs 5.28), the robustness of the combined transition window is not fully established. The paper also contains internal numerical inconsistencies that need correction. With those fixes, this could be a useful reference for model-building in heavy-ion phenomenology.

major comments (4)

[Appendix, item (1) and main text near Eq. (2)] The weak-coupling NLO input for η/s is not a derived NLO result but an assumed scaling relation: the text states 'The NLO weak coupling result is unknown; we use instead the expression η/s = 2T^3/qhat with the NLO result for qhat.' Since the central claim of a combined transition region 3 ≲ λ ≲ 14 is obtained by averaging over four quantities including η/s, this substitution is load-bearing. If the scaling relation fails at NLO in N=4 SYM, the η/s λ_c value and hence the combined window shift. Please provide evidence for the scaling relation in N=4 SYM (e.g., a check at leading-log order or against any available finite-N_c results), or present the combined window without η/s as a robustness test, or explicitly treat η/s as an uncertain input with a sensitivity analysis.
[Table I and discussion near 'For η/s this would have given...'] There is an inconsistency in the reported intersection-point value for η/s: the text says 'λ̃_c = 3.01 (instead of 9.51)', but Table I lists λ_c = 5.28 for the curvature criterion. The paper should reconcile these numbers. The large difference between the two criteria (3.01 vs 5.28, or 9.51 if that is intended) is attributed to the scaling-law substitution, but no quantitative uncertainty estimate for the combined transition region is derived from this. Please add a discussion of how much the final 3–14 window would change if the η/s λ_c were replaced by the intersection-point value.
[Appendix, items (1) and (3) and Fig. 2] The NLO weak-coupling expressions for both qhat/T^3 and η/s depend on the arbitrary UV cutoff qmax, which is chosen as qmax = 10T. This cutoff enters the Padé coefficients through f(λ,qmax) and A. The sensitivity of λ_c and the boundaries λ± to variations of qmax (e.g., 5T or 20T) is not reported. Without this, the claimed transition region, particularly its lower edge, is not fully grounded. Please provide a short sensitivity analysis for at least the two quantities that depend on qmax.
[Abstract and text introducing the method] The abstract states that next-to-leading-order calculations are available in both regimes for the considered quantities, but this is not true for η/s on the weak-coupling side, as the paper itself acknowledges in the Appendix. Please qualify the abstract and the introductory description to reflect that η/s uses the scaling-relation substitute rather than a direct NLO result.

minor comments (3)

[Various] There are several typos: 'thransition' in the summary paragraph, 'numercal' in the caption of Fig. 2, 'valus' near the intersection-point discussion, 'makes us of' in the introduction, and inconsistent use of 'Pad´ e' vs 'Padé'. Please proofread the text.
[Fig. 1 caption] The caption says 'The extrema of the curves define the border λc' but for s/s0 the extremum of the first derivative is used, while for the other quantities the extremum of the second derivative is used. Please make the caption and the accompanying text clearer about this distinction.
[References] Reference [14] is the author's own paper introducing the scaling relation. While this is appropriate, the text should more explicitly flag this as an assumption specific to this work, distinct from the published NLO results used for the other quantities.

Circularity Check

1 steps flagged · score 3.0 of 10

The η/s interpolation is built from the author's own scaling relation instead of a direct NLO weak-coupling calculation, so one of the four legs of the claimed 3 ≤ λ ≤ 14 window is grounded in a self-cited ansatz; the other three legs are independently supported.

ansatz smuggled in via citation [Main text, strategy paragraph; Appendix eqs. (1)-(2)]
"As an alternative we make use of the scaling relation for (η/s)(ˆq/T 3), which states that this combination is nearly independent of the coupling strength at weak coupling [13, 14]. Assuming that the scaling relation also applies to the weakly coupled N = 4 SYM theory, we will simply use the value of this double ratio for the LO leading-log results and the known NLO expression for hatq to set (η/s)NLO ≈ 6.173/(2π)(T 3/ˆq)NLO."

The weak-coupling input that fixes the η/s Padé approximant is not an independent NLO calculation of η/s; it is the author's own earlier scaling relation (refs. [13,14]) applied to the NLO qhat result. The η/s curve is therefore constructed from qhat rather than from η/s itself, and its λc value inherits any failure of that scaling ansatz. The paper itself acknowledges the fragility: by an intersection-point criterion η/s gives λc = 3.01 instead of the tabulated curvature-based 5.28, 'may be due in part to the fact that we have used the scaling law as a substitution for the unknown perturbative NLO result.' Since η/s is one of the four quantities averaged to obtain the headline range 3 ≤ λ ≤ 14, part of that central claim reduces to a self-cited, unverified substitution.

full rationale

The study is not definitionally circular: the λc values are extrema of Padé interpolants whose coefficients are fixed by published weak- and strong-coupling NLO expansions, so the λc values are not fitted parameters and the interpolations for s/s0, 2πT Ds, and qhat/T^3 are independently grounded in external calculations ([3,4,6,7,8,10,11,15,16]). The only load-bearing step that relies on the author's prior work is the η/s weak-coupling input, where the NLO result is unknown and is replaced by the scaling relation (η/s)(qhat/T^3) ≈ const from refs. [13,14]. That is an ansatz imported via self-citation rather than a direct computation, and the paper openly flags the resulting ambiguity in λc for η/s. Because η/s contributes one of the four legs of the averaged transition window, the central claim is partially, but not wholly, dependent on the self-cited ansatz. No uniqueness theorem is invoked, and no fitted quantity is renamed as a prediction. Score 3 reflects one load-bearing self-cited substitution in an otherwise self-contained derivation.

Assumptions & free parameters 1 free parameters · 3 assumptions · 0 invented entities

The transition-region estimate rests on the standard AdS/CFT dictionary, on the assumed scaling relation that substitutes for the missing NLO shear viscosity, and on the adequacy of the chosen Padé interpolant. The first is standard in the field; the latter two are explicitly acknowledged but not independently tested. The qmax=10T cutoff is a free choice affecting the NLO weak-coupling inputs.

free parameters (1)

qmax (UV cutoff in NLO weak-coupling qhat and η/s) = 10T (chosen, not scanned)
The NLO weak-coupling expressions for qhat and η/s depend on the arbitrary momentum cutoff qmax through ln(qmax/T); the paper sets qmax=10T 'consistent with [15]'. Different choices shift the weak-coupling input and hence the interpolated λc.

assumptions (3)

domain assumption AdS/CFT duality is valid at finite temperature, and the strong-coupling NLO results for s, η/s, D, and qhat are correct as cited.
The strong-coupling legs of all four Padé approximants are taken from the AdS/CFT literature; if the duality or those computations are wrong, the interpolation changes.
ad hoc to paper The scaling relation (η/s)(qhat/T^3) ≈ constant holds for weakly coupled N=4 SYM, allowing η/s NLO to be replaced by 6.173/(2π)(T^3/qhat)_NLO.
Stated in the text before the η/s construction; the paper acknowledges this substitution may be the cause of the discrepancy between λc values.
ad hoc to paper A generalized Padé approximant whose first coefficients match the NLO expansions is a good global representation of each quantity across all λ.
The transition point λc is read off from the curvature of this approximant; the paper calls the choice 'somewhat arbitrary' and gives an example where an intersection-point criterion changes η/s λc.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Transition from weak to strong coupling in thermal gauge theories: Lessons from ${\cal N}=4$ SYM Theory." pith.science (2026). https://pith.science/paper/EELV4OFG

@misc{pith2026250706845,
  author       = {Pith},
  title        = {Pith review of: Transition from weak to strong coupling in thermal gauge theories: Lessons from $\cal N=4$ SYM Theory},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/EELV4OFG}},
  note         = {Machine review of arXiv:2507.06845}
}

abstract

We investigate the transition between weak and strong coupling in thermal ${\cal N}=4$ supersymmetric Yang-Mills (SYM) theory as a function of 't Hooft coupling $\lambda$ for several quantities of phenomenological interest for which next-to-leading order calculations are available in both regimes. We use Pad\'e approximants to interpolate between the weak and strong coupling expansions and determine the location and width of the transition between the asymptotic regimes.

Figures

Figures reproduced from arXiv: 2507.06845 by the authors.

**Figure 2.** FIG. 2. Graphical representation of the central value [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

16 extracted references · 7 canonical work pages

[7]

J. O. Andersen, Q. Du, M. Strickland, and U. Tantary, Phys. Rev. D 105, 015006 (2022), arXiv:2111.12160 [hep- th]

work page Pith review arXiv 2022
[1]

J. M. Maldacena, Adv. Theor. Math. Phys. 2, 231 (1998), arXiv:hep-th/9711200

arXiv 1998
[2]

Aharony, S

O. Aharony, S. S. Gubser, J. M. Maldacena, H. Ooguri, and Y. Oz, Phys. Rept. 323, 183 (2000), arXiv:hep- th/9905111

arXiv 2000
[3]

S. S. Gubser, I. R. Klebanov, and A. A. Tseytlin, Nucl. Phys. B 534, 202 (1998), arXiv:hep-th/9805156

arXiv 1998
[4]

Buchel, J

A. Buchel, J. T. Liu, and A. O. Starinets, Nucl. Phys. B 707, 56 (2005), arXiv:hep-th/0406264

arXiv 2005
[5]

Finite coupling corrections to holographic predictions for hot QCD

S. Waeber, A. Sch¨ afer, A. Vuorinen, and L. G. Yaffe, JHEP 11, 087 (2015), arXiv:1509.02983 [hep-th]

work page Pith review arXiv 2015
[6]

Zhang, D.-f

Z.-q. Zhang, D.-f. Hou, and H.-c. Ren, JHEP 01, 032 (2013), arXiv:1210.5187 [hep-th]

arXiv 2013
[8]

Casalderrey-Solana and D

J. Casalderrey-Solana and D. Teaney, Phys. Rev. D 74, 085012 (2006), arXiv:hep-ph/0605199

arXiv 2006

Show all 16 references

[9]

Armesto, J

N. Armesto, J. D. Edelstein, and J. Mas, JHEP 09, 039 (2006), arXiv:hep-ph/0606245

2006 arXiv
[10]

Caron-Huot and G

S. Caron-Huot and G. D. Moore, JHEP 02, 081 (2008), arXiv:0801.2173 [hep-ph]

2008 arXiv
[11]

Ghiglieri, G

J. Ghiglieri, G. D. Moore, and D. Teaney, JHEP 03, 179 (2018), arXiv:1802.09535 [hep-ph]

2018 arXiv
[12]

S. C. Huot, S. Jeon, and G. D. Moore, Phys. Rev. Lett. 98, 172303 (2007), arXiv:hep-ph/0608062

2007 arXiv
[13]

Majumder, B

A. Majumder, B. M¨ uller, and X.-N. Wang, Phys. Rev. Lett. 99, 192301 (2007), arXiv:hep-ph/0703082

2007 arXiv
[14]

M¨ uller, Phys

B. M¨ uller, Phys. Rev. D 104, L071501 (2021), arXiv:2107.14775 [hep-ph]

2021 arXiv
[15]

Ghiglieri and H

J. Ghiglieri and H. Kim, JHEP 12, 049 (2018), arXiv:1809.01349 [hep-ph]

2018 arXiv
[16]

Buchel, Nucl

A. Buchel, Nucl. Phys. B 803, 166 (2008), arXiv:0805.2683 [hep-th]

2008 arXiv

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.

[1] [7]

J. O. Andersen, Q. Du, M. Strickland, and U. Tantary, Phys. Rev. D 105, 015006 (2022), arXiv:2111.12160 [hep- th]

work page Pith review arXiv 2022

[2] [1]

J. M. Maldacena, Adv. Theor. Math. Phys. 2, 231 (1998), arXiv:hep-th/9711200

arXiv 1998

[3] [2]

Aharony, S

O. Aharony, S. S. Gubser, J. M. Maldacena, H. Ooguri, and Y. Oz, Phys. Rept. 323, 183 (2000), arXiv:hep- th/9905111

arXiv 2000

[4] [3]

S. S. Gubser, I. R. Klebanov, and A. A. Tseytlin, Nucl. Phys. B 534, 202 (1998), arXiv:hep-th/9805156

arXiv 1998

[5] [4]

Buchel, J

A. Buchel, J. T. Liu, and A. O. Starinets, Nucl. Phys. B 707, 56 (2005), arXiv:hep-th/0406264

arXiv 2005

[6] [5]

Finite coupling corrections to holographic predictions for hot QCD

S. Waeber, A. Sch¨ afer, A. Vuorinen, and L. G. Yaffe, JHEP 11, 087 (2015), arXiv:1509.02983 [hep-th]

work page Pith review arXiv 2015

[7] [6]

Zhang, D.-f

Z.-q. Zhang, D.-f. Hou, and H.-c. Ren, JHEP 01, 032 (2013), arXiv:1210.5187 [hep-th]

arXiv 2013

[8] [8]

Casalderrey-Solana and D

J. Casalderrey-Solana and D. Teaney, Phys. Rev. D 74, 085012 (2006), arXiv:hep-ph/0605199

arXiv 2006