Bayesian inference and jet quenching

Raymond Ehlers

REVIEW 1 major objections 5 minor 14 references

Bayesian inference and jet quenching

T0 review · 1 major / 5 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 This review argues that calibrating many jet-quenching observables at once is the essential route to the quark-gluon plasma's transport properties.

desk verdict A transparent, useful proceedings review of Bayesian qhat calibrations; the field status report is accurate, but the showcase multi-observable insight rests on the very theory-uncertainty omission it admits. read the letter →

arxiv 2507.22288 v1 pith:6TA5KPDD submitted 2025-07-29 hep-ph nucl-exnucl-th

Raymond Ehlers This is my paper

classification hep-phnucl-exnucl-th

keywords Bayesianinferencejetquenchingquark-gluonplasmatransportcoefficienthadronsuppressionmodelcalibrationheavy-ioncollisions

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

These proceedings argue that Bayesian inference has become the essential tool for extracting jet transport properties of the quark-gluon plasma from jet-quenching measurements. The review's central claim is that calibrating a model against many observables at once, including inclusive hadrons, jets, D mesons, and soft-hard correlations, yields insight into the medium's dynamics and structure that single-observable comparisons cannot provide. It shows how the choice of parametrization for the transport coefficient changes the extracted posterior, and how combining hadron and jet data produces consistent but not fully reconciled results. The paper states in Section 4 that theoretical uncertainties are not yet included in the likelihood, so the quoted credible intervals should be read as conditional on that omission.

What carries the argument

The central object is the jet transport coefficient $\hat q$, the transverse momentum squared acquired by a hard parton per unit path length in the quark-gluon plasma, usually reported as $\hat q/T^3$. The machinery is Bayes' theorem, $P(\vec{\theta}|\vec{x}) \propto P(\vec{x}|\vec{\theta})P(\vec{\theta})$, which combines a prior over model parameters with a likelihood that encodes data uncertainties and their correlations to produce a posterior distribution over the parameters. The review's argument turns on comparing different parametrizations of $\hat q$: a physics-inspired form tied to hard-thermal-loop calculations and a flexible 'information field' form, a class of random functions that lets the posterior choose its own functional shape. Differential data selections are then used to see how the posterior moves when the dataset changes.

What would settle it

Re-run the hadron-only and jet-only calibrations with an explicit theory-uncertainty covariance added to the likelihood and the same flexible parametrization; if the two posterior distributions for $\hat q/T^3$ fail to converge once theory uncertainties are included, or if the combined posterior falls outside the spread of the two single-observable posteriors, the review's consistency claim would be contradicted. A simpler check is to predict an uncalibrated high-$p_T$ jet substructure observable from the combined posterior, since a large mismatch would show the calibration is missing physics rather than merely needing more data.

Watch

Extended reading notes

Core claim

The paper's central claim is that multi-observable Bayesian calibration is the most informative route to the jet transport coefficient of the quark-gluon plasma, and that the approach has matured from a proof of concept to comprehensive global analyses. It asserts that the quantity $\hat q$, the transverse momentum squared transferred to a hard parton per unit length, should be a universal property of the medium, so the extracted value should not depend on which observable is used. The review documents that combined hadron+jet calibrations extract a constrained posterior distribution for $\hat q/T^3$, that hadron-only and jet-only selections bracket the combined result within their 90% credible intervals, and that the low-$p_T$ hadron data dominate the tension because they are the most precise. It concludes that the field's next steps are to fold in theory uncertainties, improve the low-$p_T$ description, and compare multiple models under equivalent conditions.

Load-bearing premise

The load-bearing premise is that the model predictions are accurate enough that leaving theoretical uncertainty out of the likelihood does not change the results; the paper itself notes in Section 4 that the model is expected to be most uncertain in the low-$p_T$ region, where the most precise data dominate.

Editorial extensions

If this is right

Single-observable comparisons are no longer sufficient to discriminate jet-quenching models; multi-observable calibrations become the reference standard for extracting medium properties.
The extracted $\hat q/T^3$ posterior from combined hadron and jet data is consistent with other extractions within 90% credible intervals, so $\hat q$ can be treated as approximately universal while hadron-only and jet-only results bracket the combined value.
Flexible parametrizations are necessary to capture the low-temperature rise of $\hat q/T^3$ but trade away physical interpretability, so both flexible and physics-motivated parametrizations should be reported together.
Until theoretical uncertainties are added to the likelihood, quoted credible intervals should be understood as conditional on the model being exact; adding them is likely to change the intervals and may resolve the low-$p_T$ hadron tension.
Future high-precision high-$p_T$ hadron measurements and jet substructure measurements are expected to have the largest impact on narrowing the posterior.

Reading between the lines

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

One testable extension is to use the spread between hadron-only and jet-only posteriors as a proxy for model-form uncertainty and widen the reported credible intervals accordingly, giving readers an honest error bar without waiting for a full theory-uncertainty treatment.
If the review's account is right, the next set of measurements should be chosen by running sensitivity studies on the current posteriors, turning experimental design into a quantitative exercise rather than a qualitative one.
A concrete place to look for missing physics is jet substructure at high $p_T$: predicting it from the current posterior and checking where the prediction fails would localize the effects, such as coherence or nuclear shadowing, that the calibration is missing.

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

Reading this as a review, the punchline is: it's a solid, honest status report, not a new result. The author knows this; the paper explicitly says it reviews recent calibrations. The strongest part is the careful introduction of the three components of a Bayesian analysis (parametrization, data, and the inference itself) and the running example of how parametrization flexibility changes low-T behavior of qhat. The DREENA-A soft-hard calibration summary is also useful.

The paper does what it sets out to do: it synthesizes the current state of Bayesian jet-quenching calibrations, and it flags its own limitation. In particular, Section 4 states that theoretical uncertainties are not included in the likelihood, which is exactly where the stress-test concern lands. The paper's central illustration of multi-observable value is the hadron-only vs jet-only posterior separation in Fig. 5, along with the interpolation across hadron pT thresholds. The stress-test note is right: if you add a realistic theory-uncertainty term, those posteriors could become compatible, and the 'deep insight' about model deficiencies at low pT would not be established by these analyses. The paper itself acknowledges this at the end of that section, but the acknowledgment is a single sentence buried after the figure is used to make the point. The general claim that Bayesian inference is useful survives; the specific lesson drawn from Fig. 5 does not, as presented.

The selection is also JETSCAPE-centric. Most of the pedagogical examples come from the author's own collaboration's analyses. That's transparent and arguably appropriate for a proceedings, but a more balanced review would give more space to the information-field analysis and independent checks. The numerical posterior claims are quoted without error bars on the extraction procedure, but that's standard for a proceedings.

Bottom line: this is a review, and a reasonably good one. It's not a new result, so don't expect novelty. It earns its place as a community status report. I'd send it to peer review; a referee would want the theory-uncertainty caveat made more prominent, and maybe a comparison table across analyses. I'd bring it to a reading group as an entry point for grad students, and I'd cite it if I were writing my own review—but in the next 12 months I probably won't.

Referee Report

1 major / 5 minor

Summary. This proceedings article by Ehlers reviews recent applications of Bayesian inference to jet quenching in heavy-ion collisions. It introduces the key components of a Bayesian calibration—model parametrization, data selection, and the inference setup—and then summarizes three lines of work: hadron-only calibrations (JETSCAPE and information-field approaches), a soft-hard calibration with DREENA-A, and combined inclusive hadron+jet calibrations (LIDO and JETSCAPE). The paper's central thesis is that multi-observable Bayesian calibrations yield deeper insight into the quark-gluon plasma's transport properties than single-observable comparisons, and it concludes with an outlook on future directions such as theory-uncertainty treatment, multi-model comparisons, and new observables.

Significance. The paper is a review rather than original research, and its value lies in synthesizing recent literature and articulating the methodological choices that distinguish Bayesian jet-quenching analyses. It is accurate about the qualitative findings of the cited works, and it is commendably transparent about limitations, particularly the absence of theoretical uncertainties in the likelihood (Section 4). The review also gives due credit to the specific analyses and highlights concrete open questions (e.g., theory uncertainties, multi-model benchmarks). If read with the recommended caveat, it provides a useful orientation for practitioners. Its significance is moderate: it consolidates the state of the art but does not introduce new results.

major comments (1)

[Section 5 (Outlook); also Section 4 (Fig. 5)] The claim in the Outlook that 'Calibrations performed using multiple observable provide deep insights' is stronger than what the reviewed analyses strictly establish, given the repeated caveat in Section 4 that theoretical uncertainties are not incorporated in the likelihood. The specific illustration in Fig. 5—the hadron-only vs. jet-only posterior separation—could be substantially altered by a correlated theory-uncertainty term (e.g., a normalization uncertainty on low-pT R_AA), as the paper itself notes. I recommend explicitly conditioning the 'deep insights' claim on the treatment of theoretical uncertainties, or distinguishing the robust methodological lesson (that low-pT hadron data identify a region where the model is most uncertain and needs improvement) from the specific posterior shifts, which are not yet robust. This could be done with one or two sentences in the abstract and Outlook.

minor comments (5)

[General] The text contains many instances of 'di fferent' (e.g., Section 1, 'di fferent observables'; Section 2, 'di fference in data selection'), which appears to be a LaTeX hyphenation artifact; the word should be spelled 'different' throughout.
[Section 5] In the Outlook, 'Calibrations performed using multiple observable' should be 'multiple observables'.
[Section 4] The phrase 'hadron pT > 10–30 GeV/c' is ambiguous; it should refer to thresholds of 10, 20, and 30 GeV/c, as in Figure 5.
[Figure 5 caption] The caption uses '(left)' and '(Right)' with inconsistent capitalization; make both lowercase and use a consistent style.
[Section 1] The list of three key components ends with '3) and the analysis itself'; remove the 'and' for parallel structure, or rephrase as '3) the analysis itself'.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper is a review with no derivation whose claims reduce to its inputs.

full rationale

This manuscript is a proceedings-style review of Bayesian inference analyses applied to jet quenching. It introduces Bayes' theorem, discusses model selection, data selection, and analysis choices, and summarizes results from external references such as Refs. [5-9] and [12]. There is no equation in which an output is constructed from the same quantity it claims to predict, and no fitted parameter is relabeled as a prediction. The central claims about the value of multi-observable Bayesian calibrations are supported by citing published, independently performed analyses with specified likelihoods and data, rather than by using the review's own conclusion as an input. The paper explicitly flags open limitations, notably in Section 4, where it states that the current analysis does not incorporate theoretical uncertainties and that the model is expected to be most uncertain in the low-pT region; this is a correctness limitation, not a circularity. The discussion of hadron-only versus jet-only posterior tension and pT-selected interpolations is presented as an interpretation of external results, and the paper itself notes that the effect of neglecting theory uncertainties is not taken into account. No self-citation chain is load-bearing: the author's own cited work (e.g., Ref. [8]) is an external published analysis that the review merely summarizes. Therefore, the derivation chain, to the extent there is one, is self-contained and no circular step is exhibited.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The review introduces no free parameters or invented entities. It relies on standard Bayesian calculus, domain assumptions about qhat, and the representativeness of the selected analyses. The most consequential assumption is the neglect of theoretical uncertainties in the likelihood, which the review itself flags in Section 4.

assumptions (5)

standard math Bayes' theorem as the basis for posterior inference
Used in Eq. (1) to define the posterior distribution.
domain assumption The jet transport coefficient qhat characterizes energy loss in QCD matter
Assumed throughout, e.g., Section 1: 'Energy loss in QCD matter is characterized by the jet transport coefficient, qhat.'
domain assumption The qhat/T^3 scaling removes the leading T dependence
Assumed in Section 1, used for comparing extractions.
domain assumption The selected models and data sets in the reviewed analyses are representative of the field
The review selects specific analyses (JETSCAPE, information field, LIDO, DREENA-A) to draw conclusions about progress.
ad hoc to paper Theoretical uncertainties are negligible or can be ignored in the likelihood
Explicitly stated as a limitation in Section 4: 'the current analysis does not incorporate theoretical uncertainties.' The review's endorsement of the calibrated posteriors relies on this assumption being acceptable.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Bayesian inference and jet quenching." pith.science (2026). https://pith.science/paper/6TA5KPDD

@misc{pith2026250722288,
  author       = {Pith},
  title        = {Pith review of: Bayesian inference and jet quenching},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6TA5KPDD}},
  note         = {Machine review of arXiv:2507.22288}
}

read the original abstract

These proceedings review the application of Bayesian inference to high momentum transfer probes of the quark--gluon plasma (QGP). Bayesian inference techniques are introduced, highlighting critical components to consider when comparing analyses. Recent calibrations using hadron observables are described, illustrating the importance of the choice of parametrization. Additional recent analyses that characterize the impact of the inclusion of jet observables, as well as soft-hard correlations, are reviewed. Finally, lessons learned from these analyses and important questions for the future are highlighted.

Figures

Figures reproduced from arXiv: 2507.22288 by the authors.

**Figure 1.** qˆ/T 3 extracted for a physics-inspired (dotted fill) [7] and an information field [6, 9] (red) qˆ/T 3 parameterization calibrated on hadron yield modification data. Figure from Ref. [6] [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗

**Figure 2.** qˆ/T 3 extracted from differential selections of RHIC data measured at √ sNN = 200 GeV (left) and LHC data measured at √ sNN = 5.02 TeV (right) as a function of centrality selection. All calibrations are consistent at fixed √ sNN, with improved constraints for more data. Figures from Ref. [6] [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** Parameter posterior distributions from the calibration of a subset of the soft-sector parameters in the DREENA-A model for low-pT (blue) and pT-inclusive observables (green). As indicated by the red arrows, the most probable values of the distributions change with the pT-inclusive data, although the distributions are consistent within their credible intervals. Figures are adapted from Ref. [12]. The IF parameterizat… view at source ↗

Figures from the paper (2 more)

**Figure 4.** Figure 4: qˆ/T 3 posterior distributions as a function of T extracted by the calibration of the LIDO model to selected hadron and jet measurements. ˆq/T 3 is shown for E = 10 GeV (red) and E = 100 GeV (green). Figure from Ref. [5]. D-meson yield modification measurements [5]. Th…

**Figure 5.** Figure 5: (left) ˆq/T 3 posterior distribution extracted for hadron-only and jet-only observables as a function of T. (Right) ˆq/T 3 posterior distribution extracted for jet-only observables (blue), hadron-only observables (orange), and selections of hadron data for pT > 10 GeV/…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

14 extracted references · 2 canonical work pages

[1]

Bernhard, J.S

J.E. Bernhard, J.S. Moreland, S.A. Bass, Bayesian estimation of the specific shear and bulk viscosity of quark–gluon plasma, Nature Phys. 15, 1113 (2019). 10.1038/s41567- 019-0611-8

doi:10.1038/s41567- 2019
[2]

Everett et al

D. Everett et al. (JETSCAPE), Phenomenological constraints on the transport properties of QCD matter with data-driven model averaging, Phys. Rev. Lett.126, 242301 (2021), 2010.03928. 10.1103/PhysRevLett.126.242301

arXiv 2021
[3]

G. Nijs, W. van der Schee, U. Gürsoy, R. Snellings, Bayesian analysis of heavy ion collisions with the heavy ion computational framework Trajectum, Phys. Rev. C 103, 054909 (2021), 2010.15134. 10.1103/PhysRevC.103.054909

arXiv 2021
[4]

Apolinário, Y .J

L. Apolinário, Y .J. Lee, M. Winn, Heavy quarks and jets as probes of the QGP, Prog. Part. Nucl. Phys. 127, 103990 (2022), 2203.16352. 10.1016/j.ppnp.2022.103990

arXiv 2022
[5]

W. Ke, X.N. Wang, QGP modification to single inclusive jets in a calibrated transport model, JHEP 05, 041 (2021), 2010.13680. 10.1007/JHEP05(2021)041

arXiv 2021
[6]

M. Xie, W. Ke, H. Zhang, X.N. Wang, Global constraint on the jet transport coe fficient from single-hadron, dihadron, andγ-hadron spectra in high-energy heavy-ion collisions, Phys. Rev. C 109, 064917 (2024), 2208.14419. 10.1103/PhysRevC.109.064917

arXiv 2024
[7]

Cao et al

S. Cao et al. (JETSCAPE), Determining the jet transport coe fficient ˆq from inclusive hadron suppression measurements using Bayesian parameter estimation, Phys. Rev. C 104, 024905 (2021), 2102.11337. 10.1103/PhysRevC.104.024905

arXiv 2021
[8]

Ehlers et al

R. Ehlers et al. (JETSCAPE), Bayesian inference analysis of jet quenching using in- clusive jet and hadron suppression measurements, Phys. Rev. C 111, 054913 (2025), 2408.08247. 10.1103/PhysRevC.111.054913

arXiv 2025

Show all 14 references

[9]

M. Xie, W. Ke, H. Zhang, X.N. Wang, Information-field-based global Bayesian infer- ence of the jet transport coe fficient, Phys. Rev. C 108, L011901 (2023), 2206.01340. 10.1103/PhysRevC.108.L011901

2023 arXiv
[10]

Putschke et al

J.H. Putschke et al. (JETSCAPE), The JETSCAPE framework (2019), 1903.07706

2019 arXiv
[11]

Burke et al

K.M. Burke et al. (JET), Extracting the jet transport coe fficient from jet quenching in high-energy heavy-ion collisions, Phys. Rev. C 90, 014909 (2014), 1312.5003. 10.1103/PhysRevC.90.014909

2014 arXiv
[12]

M. Djordjevic et al., Bayes-DREENA: Integrated QGP Parameter Inference from High- pt and Low-pt Data, in 12th International Conference on Hard and Electromagnetic Probes of High-Energy Nuclear Collisions (2025)

2025
[13]

Karmakar, D

B. Karmakar, D. Zigic, I. Salom, J. Auvinen, P. Huovinen, M. Djordjevic, M. Djord- jevic, Constraining η/s through high-p⊥ theory and data, Phys. Rev. C 108, 044907 (2023), 2305.11318. 10.1103/PhysRevC.108.044907

2023 arXiv
[14]

Ehlers et al

R. Ehlers et al. (JETSCAPE), Measuring jet quenching with a Bayesian inference analysis of hadron and jet data by JETSCAPE, EPJ Web Conf. 296, 15009 (2024), 2401.04201. 10.1051/epjconf/202429615009

2024 arXiv

Pith tools

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

[1] [1]

Bernhard, J.S

J.E. Bernhard, J.S. Moreland, S.A. Bass, Bayesian estimation of the specific shear and bulk viscosity of quark–gluon plasma, Nature Phys. 15, 1113 (2019). 10.1038/s41567- 019-0611-8

doi:10.1038/s41567- 2019

[2] [2]

Everett et al

D. Everett et al. (JETSCAPE), Phenomenological constraints on the transport properties of QCD matter with data-driven model averaging, Phys. Rev. Lett.126, 242301 (2021), 2010.03928. 10.1103/PhysRevLett.126.242301

arXiv 2021

[3] [3]

G. Nijs, W. van der Schee, U. Gürsoy, R. Snellings, Bayesian analysis of heavy ion collisions with the heavy ion computational framework Trajectum, Phys. Rev. C 103, 054909 (2021), 2010.15134. 10.1103/PhysRevC.103.054909

arXiv 2021

[4] [4]

Apolinário, Y .J

L. Apolinário, Y .J. Lee, M. Winn, Heavy quarks and jets as probes of the QGP, Prog. Part. Nucl. Phys. 127, 103990 (2022), 2203.16352. 10.1016/j.ppnp.2022.103990

arXiv 2022

[5] [5]

W. Ke, X.N. Wang, QGP modification to single inclusive jets in a calibrated transport model, JHEP 05, 041 (2021), 2010.13680. 10.1007/JHEP05(2021)041

arXiv 2021

[6] [6]

M. Xie, W. Ke, H. Zhang, X.N. Wang, Global constraint on the jet transport coe fficient from single-hadron, dihadron, andγ-hadron spectra in high-energy heavy-ion collisions, Phys. Rev. C 109, 064917 (2024), 2208.14419. 10.1103/PhysRevC.109.064917

arXiv 2024

[7] [7]

Cao et al

S. Cao et al. (JETSCAPE), Determining the jet transport coe fficient ˆq from inclusive hadron suppression measurements using Bayesian parameter estimation, Phys. Rev. C 104, 024905 (2021), 2102.11337. 10.1103/PhysRevC.104.024905

arXiv 2021

[8] [8]

Ehlers et al

R. Ehlers et al. (JETSCAPE), Bayesian inference analysis of jet quenching using in- clusive jet and hadron suppression measurements, Phys. Rev. C 111, 054913 (2025), 2408.08247. 10.1103/PhysRevC.111.054913

arXiv 2025