{"id":"8d5943d6-1f07-49cc-8d14-534d313b8285","arxiv_id":"2512.23354","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Simultaneous afterglow+kilonova modeling of six GRBs favors BNS progenitors for four events, allows NSBH for two, and yields log M_wind = -20.23 + 0.38 log E0,J.","lead":"This paper fits afterglow and kilonova light curves simultaneously for six nearby gamma-ray bursts and infers the likely neutron-star/black-hole origin of each. It reports a new correlation between ejected wind mass and jet energy, and it is the first sample-level EM-only estimate of binary progenitor properties.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Binary-property 'confirmation' may be inherited from the EOS-based mapping, not independently measured from EM data","rationale":"The reader's weakest_assumption focuses on the POSSIS NSBH grid geometry, which is a legitimate model-grid concern affecting the two NSBH-favored events. However, the more load-bearing concern is the derivation of binary properties: the EOS-based phenomenological mapping from ejecta masses to binary parameters can imprint the very Lambda_tilde–M_Chirp trend the paper claims to confirm. This affects the credibility of the headline 'binary properties for a sizable sample' and the confirmation statement. It does not invalidate the KN identification or the BNS/NSBH model selection for the four BNS-favored events, which rest on the light-curve fits and model comparison. The proposed concrete test can settle whether the trend is data-driven or inherited. If it is inherited, the paper's binary-property claims should be downgraded to model-dependent estimates, not independent EM measurements. This is consistent with the existing CONDITIONAL verdict, so no change to the reader's verdict is needed—hence UNCHANGED.","tokens_in":43021,"tokens_out":5500,"duration_ms":56937,"concrete_test":"Perform a prior-predictive check: draw M_dyn and M_wind from the priors used in Sec 3.4, map them through the same Krüger & Foucart/Dietrich phenomenological relations and Huth EOS set, and record the implied joint distribution of M_Chirp and Lambda_tilde. If this prior-predictive distribution already shows the same anti-correlation band as Fig. 12 (or if the observed posteriors lie exactly on the prior-predictive ridge), then the claimed 'confirmation' is inherited from the mapping rather than driven by EM data. A complementary check: rerun the binary-property inference with an agnostic prior that treats M_Chirp and Lambda_tilde as independent (e.g., uniform in both over the relevant ranges) and see whether the anti-correlation persists after reweighting. If it disappears, the EM-only data do not independently confirm the relation.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim includes the first sample-level binary properties and the confirmation of the Lambda_tilde–M_Chirp anti-correlation (Sec 5.5, Fig. 12). But binary properties are not measured directly from the EM light curves: following Sec 3.4, the KN ejecta masses (M_dyn, M_wind) are mapped to component masses using phenomenological relations (Krüger & Foucart 2020; Dietrich et al. 2020) built on the Huth et al. (2022) EOS set, and Lambda_tilde is then computed from the EOS. This mapping already encodes a near-universal relation between tidal deformability and mass, so the posterior (M_Chirp, Lambda_tilde) will show an anti-correlation even if the light curves contain no independent information about the trend. The paper's claim that 'such a dependence is not prescribed during inference' (Sec 5.5) is therefore misleading: the dependence enters through the prior/phenomenological model. This affects all six events, not just the ambiguous NSBH cases, and weakens both the 'confirmation' of the Altiparmak/Magnall relations and the 'first time' binary-property results. The NSBH-grid concern raised by the reader is real but narrower; this mapping issue cuts across the entire sample.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a uniform Bayesian analysis of six merger-driven GRBs with kilonova (KN) claims, simultaneously modeling the non-thermal afterglow (afterglowpy) and thermal KN emission (POSSIS) within the NMMA framework. For each event, the authors compare afterglow-only, BNS+KN, and NSBH+KN models via Bayes factors, infer ejecta masses (M_dyn, M_wind), and then map the inferred ejecta masses to binary properties (chirp mass, tidal deformability, mass ratio) using phenomenological relations built on a fixed EOS set. The main claims are: robust KN identification in all cases except GRB 150101B; BNS progenitors favored for 160821B, 170817A, 211211A, and 230307A; a marginal preference for NSBH in 150101B and 191019A; wind mass exceeding dynamical mass; a correlation between wind mass and beaming-corrected jet energy; and a confirmation of the anti-correlation between tidal deformability and chirp mass. The paper also reports this as the first sample-level determination of merger-driven GRB progenitor properties from EM data alone.","tokens_in":43406,"tokens_out":5243,"duration_ms":49038,"significance":"If the central claims hold, the paper provides a valuable methodological demonstration: simultaneous afterglow+KN fitting on a small but homogeneous sample, including a successful blind validation on GW170817/AT2017gfo, could become the standard for EM-only progenitor classification. The use of public tools (NMMA, POSSIS, afterglowpy) is a strength, as is the explicit model comparison with Bayes factors and the inclusion of systematic uncertainties via sigma_sys. However, the binary-property results are not independent measurements: they are derived through phenomenological mass–EOS relations, so the claimed 'confirmation' of theoretical Lambda–M_chirp relations and the 'first time' narrative are weaker than presented. The main value is the ejecta-parameter and progenitor-classification exercise, with the caveats discussed below.","major_comments":[{"comment":"The claim that the inferred anti-correlation between Lambda_tilde and M_Chirp is \"not prescribed during inference\" (Sec. 5.5) is misleading. Sec. 3.4 explicitly maps M_dyn/M_wind to binary properties via the phenomenological relations of Krüger & Foucart (2020) and Dietrich et al. (2020) built on the Huth et al. (2022) EOS set. Those relations already encode a near-universal Lambda_tilde–M_Chirp dependence, so the posterior will show the anti-correlation even if the light curves contain no independent information about it. The authors should quantify the difference between the posterior and the prior predictive distribution implied by their mapping, and temper the 'confirmation' statement. Section 6.1's suggestion to prescribe this dependence 'directly in the inference framework' further indicates the dependence is not currently absent; it is mediated by the phenomenological prior.","section":"Sec. 3.4 and Sec. 5.5"},{"comment":"For GRB 191019A, the model comparison shown in Figure 1 and Table A.5 uses log(eps_e)=-0.3 and log(eps_B)=-2.0 fixed, while the parameters reported in Table 1 and the main text use freely varying eps_e/eps_B. Section 3.1 states that these parameters were freed only after inspecting the n0 posterior. This is a post-hoc model choice and introduces potential selection bias. The paper should present the full model comparison (evidence and Bayes factors) for the free-eps version, and justify that the fixed-eps model is the appropriate reference for the scientific conclusions. If the free-eps model changes the ranking, the inference is not robust.","section":"Sec. 3.1, Table 1, Table A.5, Fig. 1"},{"comment":"The claimed correlation between log M_wind and log E0,J is based on six events, with Pearson p=0.034 and Spearman p=0.019. With n=6, these p-values are fragile: they are driven by one or two points (notably GW170817), the sample is not independent of the fitting procedure (M_wind and E0 are jointly inferred with shared parameters), and no multiple-testing correction is applied to the many possible correlations examined. The authors should provide a jackknife or bootstrap test excluding each event in turn, and state the prior probability of testing this particular correlation. The abstract's 'statistically significant' wording should be softened unless these tests confirm the result.","section":"Sec. 5.4, Fig. 9"},{"comment":"The NSBH preference for GRB 150101B and GRB 191019A rests on the NSBH grid geometry in POSSIS, in which the lanthanide-rich ejecta component is fixed within Phi=30 deg and no lanthanide-poor high-latitude component exists. If true NSBH ejecta have a larger lanthanide-poor fraction, the Bayes-factor preference for the NSBH model could be an artifact of the grid, not of the data. The manuscript should discuss how the evidence changes under plausible variations of the NSBH Phi parameter (e.g., a grid with Phi=15 or 45 deg, or a lanthanide-poor outer component), and how the quoted 'slight preference' depends on this modeling assumption.","section":"Sec. 3.2, Sec. 5.5, Fig. 10"},{"comment":"All early-time data (t<0.9 d) for GRB 160821B are converted to upper limits on the grounds that a reverse/refreshed shock is present. This is a conservative treatment for a forward-shock-only model, but it discards significant information and may bias the inferred KN parameters. The impact of this choice should be tested, for example by fitting with an early reverse-shock component or at least by varying the time threshold. Given that GRB 160821B is one of the four 'BNS favored' events, the robustness of its classification to this data-handling choice should be demonstrated.","section":"Sec. 2.3"}],"minor_comments":[{"comment":"'GRB 191019A (left)' should read '(right)' — the figure shows 170817A on the left and 191019A on the right.","section":"Fig. 3 caption"},{"comment":"The text says NSBH cases have 'higher <M_Chirp>=1.3±2 M_sun'; the quoted uncertainty seems numerically wrong (Table 3 gives 2.05+0.43/-0.34 and 1.86+0.45/-0.36). Please correct the value and error.","section":"Sec. 4.2"},{"comment":"'because of its error≥0.3%' should refer to 0.3 magnitudes, not 0.3%.","section":"Sec. 2.7"},{"comment":"Typo: 'NSBN-TH' should be 'NSBH-TH'.","section":"Fig. 1 caption"},{"comment":"The caption says 'BNS-GS & BNS-TH' but the table header uses 'BNS-TH' etc. In the header row, 'NSBN-TH' appears instead of 'NSBH-TH' — please proofread the table headers.","section":"Table A.3"},{"comment":"The inclination angle is quoted as i=32.73 deg ±0.57 deg, which is consistent with the prior range Sine(0.20,0.60) rad; make clear that the posterior is very narrow, likely due to the informative prior and the high-quality data, and that the result may be prior-dominated.","section":"Sec. 4.1, GRB 170817A"}],"recommendation":"major_revision","confidential_remarks":"The paper is a solid application of existing tools to a small sample, and the validation on GW170817 is useful. The main concern is the overstatement of the binary-property results: the Lambda–M_chirp 'confirmation' is largely inherited from the phenomenological mapping and should be reframed as a consistency test of the assumed relations rather than an independent confirmation. The internal inconsistency for GRB 191019A between the model comparison (fixed eps_e/eps_B) and the reported parameters (free eps_e/eps_B) needs to be fixed before publication. I would recommend major revision and a careful re-analysis of the correlation significance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a careful, transparent application of the established NMMA/POSSIS/afterglowpy machinery to six well-observed merger-driven GRBs, anchored by a GW170817 calibration run. The simultaneous modeling is a step forward compared to the afterglow-subtraction approach favored in much of the literature, and the full per-GRB posteriors are given, which is good practice. I also buy the M_wind–E0,J correlation as a promising hint, though with six points the p-values should be treated cautiously.\n\nThe main soft spot is the binary-property claim. The posteriors for Λ̃ and M_chirp are not mapped directly from the light curves; they come from feeding the fitted ejecta masses through the Krüger/Foucart and Dietrich phenomenology on the Huth EOS set. That mapping already encodes a tight Λ̃–M relation. So the anti-correlation in Fig. 12 is largely a projection of the assumed EOS/phenomenology, not something the light curves themselves demand. The authors say the dependence is 'not prescribed,' which is technically true, but the mapping effectively does the prescribing. Their own future-perspectives section says they plan to prescribe it directly next time. So the 'confirmation' should be labeled a consistency check, and the 'first time' binary properties have to be read as model-dependent outputs of the EOS mapping, not independent EM measurements. This weakens one of the headline results.\n\nOther soft spots are minor: the abstract overclaims that every GRB shows a KN when the text admits 150101B is inconclusive; the eps_e/eps_B loosening for 191019A after a low-density finding is honest but should be flagged as exploratory; and converting the early 160821B data to upper limits throws away information, though it is defensible given reverse-shock contamination.\n\nThe NSBH-grid geometry issue is real but narrow, and the authors themselves say the NSBH preference is 'slight' and BNS remains viable for both events.\n\nOverall: worth a serious referee. The referee should ask the authors to reframe the binary-property discussion, separate the cleaner KN-ejecta results from the model-dependent progenitor properties, and either fix the six-point correlation statistics or add an explicit caveat.","headline":"Solid joint afterglow+KN inference for six GRBs, but the binary-property 'confirmation' is largely inherited from the EOS-based mapping, not an independent EM measurement.","tokens_in":43912,"tokens_out":3798,"would_cite":true,"duration_ms":38849,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"By fitting the afterglow and kilonova of six gamma-ray bursts in a single Bayesian model, this paper claims that electromagnetic data alone can identify whether the progenitor was a binary neutron star or a neutron star–black hole and recov","keywords":["gamma-ray bursts","kilonova","neutron star mergers","NSBH mergers","afterglow","Bayesian inference","tidal deformability","chirp mass"],"falsifier":"Re-run the analysis for GRB 150101B and GRB 191019A with an NSBH kilonova grid that permits lanthanide-poor ejecta at high latitudes; if the Bayes factor no longer prefers NSBH, the classification is a grid artifact. A future gravitational-wave-detected NSBH merger that shows a blue, lanthanide-poor kilonova would likewise contradict the fixed-geometry assumption.","tokens_in":42911,"feed_emoji":"💥","tokens_out":8455,"duration_ms":70822,"temperature":0.7,"pith_summary":"The paper aims to show that electromagnetic observations alone—without gravitational waves—can reveal what kind of compact-object merger produced a gamma-ray burst. It models the non-thermal afterglow and the thermal kilonova emission together in a single Bayesian fit for six nearby GRBs with claimed kilonovae, using GW170817/AT2017gfo as a calibrator. It finds that a kilonova is clearly present in five of the six events, that four events favor a binary neutron star progenitor, and that two slightly favor a neutron star–black hole merger. It then infers, for the first time from EM data alone, sample-level binary properties—chirp mass, tidal deformability, and mass ratio—and reports that the disk wind ejecta typically outweigh the dynamical ejecta by about a factor of two. If right, this turns existing GRB data into a population-level probe of neutron star mergers and their equation of state.","feed_headline":"Kilonova fits classify six gamma-ray bursts as neutron-star mergers","feed_subtitle":"Joint afterglow+kilonova fits give binary masses and deformability without gravitational waves","key_machinery":"The load-bearing tool is a joint Bayesian model that couples a forward-shock afterglow light-curve model (with Gaussian or top-hat jet geometries) to a multi-component kilonova model built from two ejecta components—fast dynamical ejecta and a slower disk wind—with a free half-opening angle for the lanthanide-rich region. The same viewing angle is shared by both emission components, so the fit can separate thermal kilonova light from non-thermal afterglow without subtracting one from the other. The inferred ejecta masses are then mapped to binary properties (chirp mass, tidal deformability, mass ratio) through equation-of-state-dependent phenomenological relations.","core_discovery":"The paper's central claim is that a single Bayesian fit to both the non-thermal afterglow and the thermal kilonova can simultaneously identify the merger type and recover binary properties from electromagnetic data alone. Applied to six nearby GRBs with claimed kilonovae—using GW170817/AT2017gfo as a calibrator—the joint fit clearly identifies a kilonova in five events (with GRB 150101B ambiguous), favors a binary neutron star progenitor for GRB 160821B, GRB 170817A, GRB 211211A, and GRB 230307A, and leans slightly toward a neutron star–black hole progenitor for GRB 150101B and GRB 191019A. The fit also yields, for the first time from EM data alone, sample-level binary parameters: wind eject","pith_inferences":["If the M_wind–E0,J correlation survives a larger sample, kilonova brightness could become a proxy for jet energy in GRBs lacking X-ray afterglows, and the slope could be compared with disk-mass prescriptions to measure accretion-to-jet efficiency.","The paper's separation of short and long merger-driven GRBs in the mass-ratio versus dynamical-mass plane hints that burst duration may trace the amount of fallback accretion; this could be tested by measuring q and M_dyn for more long-duration merger GRBs.","A straightforward robustness check would be to rerun the NSBH fits with a grid that includes lanthanide-poor polar ejecta; the paper's classifications of GRB 150101B and GRB 191019A currently rest on the fixed 30-degree lanthanide-rich geometry.","The method's calibrated performance on GW170817 suggests it could one day be used to rank electromagnetic candidates for follow-up even before a gravitational-wave alert is available, though the paper does not discuss this."],"forward_implications":["Kilonova emission is present in essentially all well-observed nearby merger-driven GRBs; the one ambiguous case is GRB 150101B, where the data cannot confirm or exclude the kilonova.","Four of the six events (GRB 160821B, GRB 170817A/AT2017gfo, GRB 211211A, GRB 230307A) favor a binary neutron star progenitor, while GRB 150101B and GRB 191019A slightly favor a neutron star–black hole system with BNS still viable.","Across the sample, the disk wind ejecta (median 0.027 solar masses) dominate over the dynamical ejecta (median 0.012 solar masses), so post-merger winds carry most of the mass that powers the kilonova.","The wind mass scales with beaming-corrected jet kinetic energy as log M_wind = -20.23 + 0.38 log E0,J, a statistically significant correlation reported here for the first time.","The inferred chirp mass and tidal deformability follow the expected inverse trend—lower chirp mass means higher tidal deformability—and for GW170817/AT2017gfo the EM-only values match the gravitational-wave measurements, validating the approach."],"fun_headline_variants":["Joint afterglow+kilonova fits type six GRBs as neutron-star mergers","EM-only fits classify GRBs and extract binary properties","Kilonova+afterglow modeling reveals neutron-star merger origins","Six gamma-ray bursts traced to neutron-star mergers via joint fits","Single Bayesian fit to light and kilonova pinpoints merger type"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The NSBH classifications for GRB 150101B and GRB 191019A rest on the grid assumption that NSBH ejecta have no lanthanide-poor polar component; if real NSBH ejecta can be lanthanide-poor there, the slight NSBH preferences could be artifacts of the model grid.","fun_headline_variants_meta":{"raw":{"variants":["Joint afterglow+kilonova fits type six GRBs as neutron-star mergers","EM-only fits classify GRBs and extract binary properties","Kilonova+afterglow modeling reveals neutron-star merger origins","Six gamma-ray bursts traced to neutron-star mergers via joint fits","Single Bayesian fit to light and kilonova pinpoints merger type"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000249,"raw_usage":{"total_tokens":1503,"prompt_tokens":974,"completion_tokens":529,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":718,"completion_tokens_details":{"reasoning_tokens":439}},"tokens_in":718,"tokens_out":529,"duration_ms":5537,"temperature":1.0,"reasoning_tokens":439,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T13:39:24.276854+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-run the analysis for GRB 150101B and GRB 191019A with an NSBH kilonova grid that permits lanthanide-poor ejecta at high latitudes; if the Bayes factor no longer prefers NSBH, the classification is a grid artifact. A future gravitational-wave-detected NSBH merger that shows a blue, lanthanide-poor kilonova would likewise contradict the fixed-geometry assumption.","supporting_citations":[],"review_version":1}