{"id":"450d4d49-2f35-43c9-89bf-4a2e298436d4","arxiv_id":"2509.03289","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Bayesian-calibrated hydrodynamic simulations predict that thermal photons and dileptons become relatively stronger at low collision energies and reveal sensitivity to early-stage dynamics.","lead":"This paper simulates the light and particle-pair signals emitted by the quark-gluon plasma produced in heavy-ion collisions at different energies. At low collision energies, the simulated thermal glow is predicted to dominate the prompt radiation even at high momentum, which could make the plasma easier to observe.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claimed high-pT thermal-over-prompt excess at low collision energies rests on an unvalidated NLO pQCD prompt photon prediction at large Bjorken-x; if the true prompt yield is larger, the excess disappears.","rationale":"The reader's weakest assumption is exactly the most load-bearing concern: the thermal-over-prompt excess at high pT in low-energy collisions depends on the steep decline of INCNLO prompt photons at large Bjorken-x, which is neither verified against low-energy data nor accompanied by an uncertainty estimate in the right panel of Fig. 1. I considered the alternative candidate that the 'sensitivity to hotspot size' is not isolated because the Posterior 1 vs Posterior 2 comparison changes multiple calibrated parameters simultaneously; that is a real limitation, but it affects only the second half of the concluding claim and is consistent with the paper's cautious wording. The prompt-photon issue directly threatens the paper's headline prediction. The paper does provide independent support elsewhere: the 200 GeV left-panel comparison with STAR/PHENIX validates the framework at high energy, and the pre-equilibrium comparison in Fig. 2 left is a controlled model variation. However, those do not insulate the low-energy high-pT extrapolation. Because the concern is specific and testable but does not invalidate the thermal emission calculations or the dilepton sensitivity to pre-equilibrium dynamics, the appropriate verdict remains conditional, same as the reader's.","tokens_in":3905,"tokens_out":4479,"duration_ms":44925,"concrete_test":"Recompute the right panel of Fig. 1 using an independent NLO pQCD prompt-photon generator (e.g., JETPHOX) with the same PDFs and scale choices as INCNLO, and also run a version of INCNLO with intrinsic-kT smearing or threshold resummation at sqrt(s) = 14.6, 17.3, and 19.6 GeV. Then compare the thermal-to-prompt ratio at pT = 3 and 4 GeV. If either alternative raises the prompt yield enough to push the ratio below unity, the central claim of thermal dominance at high pT in low-energy collisions is not robust. Additionally, compare the prompt-only prediction against any available direct-photon data or upper limits at SPS energies in the same pT range.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central novelty is the right panel of Fig. 1: at sqrt(s) = 14.6-17.3 GeV and pT > 2.5 GeV, the calculated thermal photon yield exceeds the prompt photon yield. The paper attributes this to the statement that 'our calculations of prompt photons fall steeply at large Bjorken-x.' That steep fall is the load-bearing assumption. At these collision energies and pT ~ 3-4 GeV, x_T = 2 pT / sqrt(s) ~ 0.4-0.55, a regime where NLO pQCD is known to be less reliable without resummation or intrinsic transverse momentum effects. The left panel validates the direct photon calculation only at sqrt(s) = 200 GeV, where prompt photons dominate at high pT and are compared to STAR and PHENIX data; no low-energy high-pT data are shown. The quoted scale-variation uncertainty (0.5 pT to 2 pT) is discussed for the 200 GeV case, but the low-energy thermal-to-prompt ratio is presented without a corresponding uncertainty band. If the true prompt background is larger than INCNLO predicts -- for example because threshold logarithms or higher-order corrections enhance near-threshold production -- the claimed thermal dominance at high pT would be reduced or eliminated, directly undermining the conclusion that 'thermal emission may be measurable.' This is a correctness risk in the prompt background calculation, not in the thermal emission calculation itself.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This proceedings-style paper presents electromagnetic radiation predictions from the iEBE-MUSIC hydrodynamic framework with finite-baryon-density Bayesian-calibrated initial conditions. The authors compute thermal photon and thermal dilepton yields by convolving in-medium emission rates with posterior-sampled hydrodynamic backgrounds, adding prompt photons from INCNLO. They report three main results: (i) at 200 GeV the direct photon spectrum is consistent with STAR and PHENIX data; (ii) at low collision energies (14.6-17.3 GeV) the calculated thermal photon yield exceeds the prompt yield even at high transverse momentum, which they attribute to the steep fall of NLO pQCD prompt photons at large Bjorken-x; and (iii) intermediate-mass dilepton spectra from In+In collisions at 17.3 GeV are sensitive to the pre-equilibrium initialization time and to the initial-state hotspot size encoded in different Bayesian posterior sets.","tokens_in":4329,"tokens_out":5920,"duration_ms":55558,"significance":"If the central claims hold, the paper identifies a potentially observable window for thermal photon emission at RHIC-BES and SPS energies, and it demonstrates that intermediate-mass dileptons can discriminate among early-time scenarios, which would motivate future Bayesian analyses that include electromagnetic probes. A notable strength is the absence of circularity: no electromagnetic data are used in the calibration, and the thermal emission rates are independent inputs. The 200 GeV direct photon comparison with STAR/PHENIX provides a useful sanity check. However, the main predictive claim about low-energy high-pT thermal photons rests on an unvalidated prompt-photon background calculation, and the dilepton sensitivity claims are not quantified with uncertainties or goodness-of-fit measures. The potential impact is real but conditional on addressing these correctness risks.","major_comments":[{"comment":"The central claim that thermal photons dominate over prompt photons at high pT for low collision energies rests entirely on the INCNLO prediction that 'our calculations of prompt photons fall steeply at large Bjorken-x.' At sqrt(sNN)=14.6-17.3 GeV and pT around 3-4 GeV, x_T = 2pT/sqrt(s) is approximately 0.4-0.55, a regime where NLO pQCD without threshold resummation or intrinsic transverse momentum is known to be less reliable. The left panel validates the prompt calculation only at 200 GeV, where x_T is much smaller for the same pT range, and the quoted scale-variation uncertainty (0.5pT to 2pT) is not shown for the low-energy ratio. Please provide the scale-variation band for the thermal-to-prompt ratio at low energies, compare the prompt predictions with existing low-energy direct photon measurements (e.g., from the SPS program), or use a resummed calculation, and state how the claimed excess depends on this background. Without this, the high-pT thermal-over-prompt observation is not established.","section":"Section 3, Fig. 1 right panel"},{"comment":"The statements that instantaneous initialization at tau0=0.5 fm/c gives 'an even better description of the NA60 data' and that Posterior 1 gives 'a larger yield in the IMR' are not supported by any quantitative measure. The plotted model curves have no uncertainty bands from the posterior sampling or from the thermal emission rates, and no chi-square, likelihood, or data-to-model ratio is reported for the different scenarios. Since the sensitivity of dileptons to the pre-equilibrium stage and to the hotspot size is a central conclusion, please provide a goodness-of-fit comparison or at least show posterior/statistical uncertainties on the model curves so that the reader can assess whether the differences among the scenarios are significant.","section":"Section 3, Fig. 2"},{"comment":"Only thermal dileptons from the QGP phase are included in the calculation, yet the text states that the emitted dilepton spectrum 'is compatible with NA60 data' without specifying the invariant-mass range displayed in Fig. 2. If the plotted comparison includes the low-mass region (LMR), hadronic thermal dilepton sources would be non-negligible and their omission would make the comparison misleading. Please clarify the mass range shown and either justify the neglect of hadronic thermal dileptons or include them. This issue also affects the comparison of Posterior 1 and Posterior 2 in the LMR, where the two sets are claimed to give similar yields.","section":"Section 2 and Section 3"}],"minor_comments":[{"comment":"The text states that the prompt photon uncertainty is estimated by scale variation from 0.5pT to 2.0pT, but the right panel of Fig. 1 shows no such uncertainty band. State explicitly that only central values are shown in that panel.","section":"Section 3, Fig. 1"},{"comment":"The caption of Fig. 2 uses 'dNch/deta >30' but this quantity is not defined in the text; please define it or give a reference for the selection.","section":"Section 3, Fig. 2"},{"comment":"The phrase 'posterior-sampled backgrounds' is not elaborated; a brief description of how many posterior samples were used and how they were propagated would help the reader understand the robustness of the results.","section":"Abstract and Section 2"},{"comment":"The error bars on the NA60 data are not described in terms of statistical versus systematic uncertainties; please specify what is shown.","section":"Section 3, Fig. 2"},{"comment":"The sentence 'Posterior 1 [4] is the default configuration in this proceeding' is awkward; 'proceeding' appears to refer to this manuscript, and the wording should be revised for clarity.","section":"Section 3"}],"recommendation":"major_revision","confidential_remarks":"The paper is a proceedings-style contribution and the claims are forward-looking rather than definitive. The main correctness risk is the unvalidated low-energy prompt-photon background, which can be addressed with additional uncertainty bands and comparisons to existing SPS data. The dilepton sensitivity claims would also benefit from quantified comparisons. With these additions, the paper would be suitable for the proceedings; in its current form, the central observational claim is not yet sufficiently supported."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper gives the heavy-ion community new calculations of thermal photons and dileptons at finite baryon density, using a Bayesian-calibrated iEBE-MUSIC framework. That is genuinely useful: it extends a mature tool to the BES/SPS regime, folds in independent hadronic calibrations, and shows clean sensitivity studies for pre-equilibrium initialization and posterior choices. No EM data are used in the fit, so the circularity burden is low, and the 200 GeV direct-photon comparison with STAR/PHENIX provides a sanity check that the framework still works in a known regime. The energy scan of the thermal-to-prompt ratio is the kind of plot experimentalists will want to see.\n\nThe soft spot is exactly where the stress-test note points. The claim that thermal photons dominate at high pT in 14.6–17.3 GeV collisions rests on INCNLO's steep fall at large Bjorken-x. That is a load-bearing part of the central conclusion, but the paper gives no validation of prompt production in that kinematic region—the left-panel comparison is at 200 GeV, where x_T is far smaller—and the right panel has no uncertainty band on the prompt component. The scale-variation discussion for 200 GeV cannot simply be assumed to apply at x_T ~ 0.4–0.55, where threshold logarithms and intrinsic-kt effects are known to matter. So the 'thermal may be measurable' statement should be read as conditional on the prompt background being right, not as a robust prediction. That does not sink the paper; it does mean the headline result is a prompt to investigate, not a settled finding.\n\nOther concerns are minor and typical for a proceedings. The 'even better description' of NA60 data is not backed by a goodness-of-fit metric, and no code or data artifacts are provided. Neither is disqualifying here—the plots speak reasonably clearly—but they do limit how much weight a referee can put on fine distinctions.\n\nOverall, this is a competent, honest contribution from people who know the framework. The math and rate inputs look fine, the qualitative statements align with the figures, and the authors are transparent about model choices. It deserves serious referee attention, but the referee should push for a quantitative treatment of the prompt-photon uncertainty at low energies, ideally a resummed or threshold-improved calculation, before the main claim is taken as strong evidence.","headline":"A useful finite-baryon extension of the iEBE-MUSIC EM calculations, but the marquee high-pT thermal-over-prompt result leans on an unvalidated NLO prompt-photon prediction at large x.","tokens_in":4753,"tokens_out":1726,"would_cite":true,"duration_ms":18401,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"At low collision energies, thermal photons from the quark-gluon plasma can outnumber prompt photons even at high transverse momentum, and dilepton spectra can reveal the plasma's earliest moments.","keywords":["thermal photons","thermal dileptons","quark-gluon plasma","finite baryon density","heavy-ion collisions","beam energy scan","Bayesian calibration","pre-equilibrium dynamics"],"falsifier":"Measure prompt photon production in proton-proton or proton-nucleus collisions at the same low nucleon-nucleon energies up to high pT and compare with the NLO pQCD prediction; if the measured prompt yield substantially exceeds the calculation in the region where the paper predicts thermal dominance (pT above about 2.5 GeV at 14.6 GeV), the claimed thermal excess in Au+Au or Pb+Pb would be an artifact of an underestimated background.","tokens_in":3701,"feed_emoji":"⚛️","tokens_out":9392,"duration_ms":81061,"temperature":0.7,"pith_summary":"This paper tries to establish that electromagnetic radiation—photons and dileptons—from the quark-gluon plasma can be isolated from competing sources at the relatively low collision energies explored by the beam energy scan. Using a hydrodynamical model calibrated to hadron data through Bayesian inference, the authors compute thermal photon emission at finite baryon density and find that at low collision energies thermal photons overtake prompt photons even at high transverse momentum, making the plasma glow measurable. They also show that the intermediate-mass dilepton spectrum is sensitive to how soon the plasma forms and to the size of initial hot spots, so dileptons can serve as a probe of pre-equilibrium dynamics. If correct, these results would extend electromagnetic probes of the QGP to the high-baryon-density region of the QCD phase diagram and motivate including such observables in future model calibrations.","feed_headline":"Thermal photons may dominate high-momentum light at low energies","feed_subtitle":"Bayesian-calibrated hydrodynamics suggests the quark-gluon plasma's own glow could be measured in low-energy collisions.","key_machinery":"The machinery is the multistage hydrodynamic model used here: three-dimensional dynamical initial conditions deposit energy and momentum gradually between proper times 0.5 and 1.5 fm/c, a viscous relativistic hydrodynamic code evolves the medium, a hadronic afterburner completes the evolution, and Bayesian inference fixes the parameters using hadronic data from the RHIC beam energy scan. Thermal photon and dilepton yields are then obtained by convolving in-medium emission rates—partonic and hadronic for photons, QGP-phase for dileptons, with the switch at 160 MeV—with local fluid-cell temperature, flow velocity, and baryon chemical potential, while prompt photons come from an NLO perturbative QCD calculation. This setup is what lets the authors compute the thermal-to-prompt photon ratio and the response of dilepton spectra to different initialization procedures.","core_discovery":"The central result is that the balance between thermal and prompt photon production shifts with collision energy: while at 200 GeV prompt photons dominate for transverse momenta above about 2.5 GeV, at energies around 14.6–17.3 GeV the thermal component exceeds the prompt one even at high pT, because the next-to-leading-order perturbative QCD prompt yield falls steeply at large Bjorken-x. For dileptons, the paper finds that the intermediate-mass-region excess is enhanced when the plasma is initialized early (at 0.5 fm/c) rather than late (1.5 fm/c), and is larger for the posterior parameter set with smaller initial hotspots (0.1 fm) than for one with larger hotspots (0.3 fm). These comparisons with STAR and PHENIX photon data and with NA60 dilepton data indicate that electromagnetic observables are sensitive to the initial state and pre-equilibrium stage at finite baryon density.","pith_inferences":["The steep fall of the prompt-photon calculation at large Bjorken-x implies that the thermal excess should grow further at collision energies below 14.6 GeV, a trend this paper does not itself compute.","Because dilepton yields depend on both initialization time and hotspot size, combining photon and dilepton measurements at the same energy could break degeneracies between pre-equilibrium timing and initial-state geometry in future Bayesian fits.","If the high-pT thermal excess is confirmed, high-pT photon analyses at low collision energies would need to subtract a plasma component rather than assume all high-pT photons are prompt."],"forward_implications":["At collision energies near 14.6–17.3 GeV, thermal photons outnumber prompt photons even for pT above 2.5 GeV, so the thermal component is not confined to low pT and its identification becomes less ambiguous.","The intermediate-mass dilepton excess is larger when the quark-gluon plasma is formed early (0.5 fm/c) than when formation is delayed to 1.5 fm/c, making the IMR a probe of pre-equilibrium dynamics.","Posterior parameter sets with smaller initial hotspots (0.1 fm) yield more intermediate-mass dileptons than sets with larger hotspots (0.3 fm), so dilepton spectra can distinguish initial-state geometry choices.","The demonstrated sensitivity of photons and dileptons motivates including electromagnetic observables in future Bayesian calibrations to tighten constraints on QGP properties at finite baryon density."],"supporting_citations":[{"why":"Supplies the 3D dynamical initial conditions used to create the hydrodynamic backgrounds.","marker":"[1]"},{"why":"Provides the viscous relativistic hydrodynamic evolution code used to simulate the medium.","marker":"[2]"},{"why":"Gives the thermal photon emission rates, including viscosity corrections, for both partonic and hadronic phases.","marker":"[3]"},{"why":"Supplies the default Bayesian posterior parameter set calibrated to RHIC-BES hadronic data.","marker":"[4]"},{"why":"Provides the thermal dilepton emission rate from the QGP phase used for the dilepton spectra.","marker":"[5]"},{"why":"Provides the NLO pQCD prompt photon calculation that sets the background against which the thermal excess is judged.","marker":"[6]"},{"why":"Supplies the STAR direct photon data used to validate the 200 GeV calculation.","marker":"[7]"},{"why":"Describes the dynamical initialization procedure whose timing sensitivity is tested in the dilepton spectra.","marker":"[9]"},{"why":"Supplies the NA60 dilepton excess data used to compare the In+In predictions.","marker":"[10]"},{"why":"Provides an alternative Bayesian posterior with a larger hotspot size, used to assess sensitivity to initial-state geometry.","marker":"[11]"}],"fun_headline_variants":["Low-energy collisions amplify the plasma's thermal glow","Thermal photons dominate prompt light at low collision energies","Dilepton spectra probe the plasma's early initialization","Bayesian hydrodynamics shifts photon balance at low energies"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The high-pT thermal excess at low collision energies rests on the NLO perturbative QCD prompt-photon calculation remaining accurate at the large Bjorken-x values those collisions reach; if the true prompt background is larger there, the claimed thermal dominance disappears.","fun_headline_variants_meta":{"raw":{"variants":["Low-energy collisions amplify the plasma's thermal glow","Thermal photons dominate prompt light at low collision energies","Dilepton spectra probe the plasma's early initialization","Bayesian hydrodynamics shifts photon balance at low energies"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00017,"raw_usage":{"total_tokens":1211,"prompt_tokens":834,"completion_tokens":377,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":450,"completion_tokens_details":{"reasoning_tokens":315}},"tokens_in":450,"tokens_out":377,"duration_ms":3952,"temperature":1.0,"reasoning_tokens":315,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T16:30:54.264815+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure prompt photon production in proton-proton or proton-nucleus collisions at the same low nucleon-nucleon energies up to high pT and compare with the NLO pQCD prediction; if the measured prompt yield substantially exceeds the calculation in the region where the paper predicts thermal dominance (pT above about 2.5 GeV at 14.6 GeV), the claimed thermal excess in Au+Au or Pb+Pb would be an artifact of an underestimated background.","supporting_citations":[],"review_version":2}