{"id":"2ac1a2ed-a012-4005-b536-7c42b2ce0b0c","arxiv_id":"1908.11408","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"HAWC's new energy estimator yields a Crab spectrum beyond 100 TeV and 95% confidence lower limits on the cutoff energy above 200 TeV for the hardest Galactic source.","lead":"The HAWC observatory applies a new neural-network energy reconstruction to measure the Crab Nebula spectrum beyond 100 TeV and to set lower limits on how far the spectra of seven Galactic sources extend. The highest limit implies photons above 200 TeV at 95% confidence from the source 2HWC J1825-134.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Energy-scale systematics dominate the 200 TeV claim, but no internal inconsistency is demonstrated.","rationale":"The reader's weakest_assumption correctly identifies the energy-estimator calibration as the load-bearing point, and I agree with the CONDITIONAL verdict. My reading of the paper supports this: the central claim is explicitly stated in the abstract, and its operational support is Table 1, where every number depends on the NN energy estimator calibrated to simulation. The paper's own Section 4 footnote defers systematic errors to a future publication, and Section 3 explicitly notes that comparisons with other experiments do not consider energy-scale systematics. These are in-text acknowledgments of the limitation, so flagging them is not importing outside criticism. The concern is not that the analysis is wrong, but that the headline claim is premature until the energy-scale systematics are quantified. The specific number of 15% is a reasonable conjecture based on the resolution shown in Fig. 1(a) and on the typical HAWC systematic error estimates; the concrete test would settle whether that shift actually moves the headline claim. I do not see a separate internal inconsistency: the hard-cutoff likelihood method is sound, the p-values are consistent with no cutoff preference, and the stated conclusion is properly qualified in Section 5 as 'evidence for 100 TeV photons' rather than a measured cutoff. Thus the verdict should remain CONDITIONAL, not REJECT or UNVERDICTED. The paper is a clear, honest conference proceedings; the only adjustment needed is either a quantitative systematic error treatment or a softened abstract before the result is cited as a definitive measurement.","tokens_in":4958,"tokens_out":1829,"duration_ms":15089,"concrete_test":"Run the published analysis reduced with a single global rescaling of the energy estimator (e.g., multiply E_hat by 0.85 and by 1.15) and recompute the Table 1 Ec(95%) limits for 2HWC J1825-134, 2HWC J1908+063, and the Crab. If both rescaled values of Ec(95%) remain above 200 TeV for at least one source, the headline claim survives an energy-scale miscalibration of 15%; if the 0.85 rescaling drops all three below 200 TeV, the headline claim is not robust without a quantitative systematic error budget. A complementary check is to compare HAWC's energy-scale-sensitive spectrum with the published LHAASO Crab measurement in the 20-100 TeV overlap region, where a slope-matched flux offset would directly quantify the absolute energy-scale bias.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim rests on the neural-network energy estimator being unbiased in an absolute sense: the abstract states photons above 200 TeV are observed at 95% confidence, which is operationally supported by Table 1 only if the simulated energy response is correct. Section 4 explicitly defers systematic errors to a future publication, and the paper itself concedes (Section 3) that cross-experiment comparisons are limited by unquantified energy-scale differences. The Crab spectrum figure shows systematic bands, but those bands are carried over from Ref. [1] and are not propagated into the Ec lower limits of Table 1. A 15% shift in energy scale would move Ec(95%) by a comparable fraction: for Crab, 152 TeV could move to roughly 130 or 175 TeV, and for 2HWC J1825-134 the 253 TeV limit could drop below 215 TeV. That would weaken but not necessarily destroy the headline conclusion, since 2HWC J1825-134 would likely remain above 200 TeV under a downward shift. The claim 'photons above 200 TeV at 95% confidence' is therefore conditional on an absolute energy-scale calibration that is not yet demonstrated in this proceedings paper. This concern is not an accusation of error; the estimator performance shown in Fig. 1(a) is based on simulations, and simulation-to-data discrepancies in the PMT charge response are the classic source of such bias. The analysis prescription is otherwise internally coherent: the hard-cutoff profile likelihood (Section 4) does not depend on the spectral shape above Ec, and the Table 1 p-values consistently show no preference for a cutoff.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This ICRC2019 proceedings paper reports HAWC's neural-network (NN) energy reconstruction for gamma rays and applies it to two analyses. First, it presents the HAWC Crab Nebula spectrum from roughly 1 TeV to beyond 100 TeV, comparing fits with previous HAWC results, other VHE observatories, and inverse-Compton models. Second, it derives lower limits on a hard spectral cutoff energy Ec for the Crab and six other Galactic sources using a profile-likelihood method that accounts for mis-reconstructed events from lower energies. The abstract's headline claim is that HAWC 'observed photons above 200 TeV at 95% confidence'; in the body this is operationalized as 95% CL lower limits on Ec in Table 1: 253 TeV for 2HWC J1825-134, 213 TeV for 2HWC J1908+063, and 152 TeV for the Crab. The paper also states that at 95% CL, five sources have Ec above 100 TeV, and at the 3-sigma level two do.","tokens_in":5242,"tokens_out":5990,"duration_ms":61699,"significance":"If the result holds, HAWC would provide the strongest current evidence that at least two Galactic sources emit gamma rays beyond 200 TeV, and the paper demonstrates the utility of the NN energy estimator for extending the Crab spectrum beyond 100 TeV. The analysis has clear strengths: an explicit forward-folding likelihood procedure, p-values for cutoff preference showing no significant preference for a cutoff, a multi-source comparison, and a direct comparison with theoretical inverse-Compton predictions. However, the claims are explicitly preliminary: systematic uncertainties are deferred to a future publication, the abstract overstates what the Table 1 limits show, and the hard-cutoff model is acknowledged to be non-physical. The central astrophysical conclusion is therefore credible but conditional on an absolute energy-scale calibration that is not demonstrated in this proceedings contribution.","major_comments":[{"comment":"The abstract's sentence 'We have observed photons above 200 TeV at 95% confidence' is not supported by the body's statistic. The analysis produces lower limits on a hard cutoff energy Ec under a fitted spectral model; Table 1 gives Ec(95%) values, and the highest estimated-energy bins discussed in Sec. 4 are only subdivided up to 177 TeV. No event-level observation of photons above 200 TeV is presented. The sentence should be rephrased to state that two sources have a 95% CL lower limit on a hard cutoff energy above 200 TeV, with the model-dependence stated explicitly.","section":"Abstract and Sec. 4, Table 1"},{"comment":"Systematic uncertainties are explicitly deferred ('A future publication will also consider the effects of systematic errors'), and the systematic bands shown in Fig. 1(b) are carried from Ref. [1] and are not propagated into the Ec limits in Table 1. Because the NN energy estimator is calibrated on simulated air showers, an unquantified energy-scale bias of order 15% would shift the Ec limits by a comparable fraction (for example, the Crab limit could drop from 152 TeV to about 130 TeV, and 2HWC J1825-134 from 253 TeV to about 215 TeV). The 200 TeV headline therefore depends on an absolute energy-scale calibration that this paper does not establish. The limits should either include a systematic term or be explicitly labeled as statistical-only.","section":"Sec. 4, footnote 1; Table 1"},{"comment":"The hard-cutoff model is acknowledged as 'not astrophysically motivated,' but the paper treats the resulting Ec limits as evidence about maximum photon energies. For the six non-Crab sources the adopted spectral model is a power law with exponential cutoff, so the hard-cutoff parameter Ec is a test statistic rather than the physical cutoff energy of the source. The authors should state explicitly that the quoted limits exclude a hard cutoff at Ec but do not directly constrain a gradual cutoff or spectral softening, and should avoid wording that equates a lower limit on Ec with a direct observation of photons above that energy.","section":"Sec. 4"}],"minor_comments":[{"comment":"There is a typo in the description of the Fraschetti and Pohl model: 'synchrotron self-Compton and and cosmic microwave background' should read 'synchrotron self-Compton and cosmic microwave background.'","section":"Sec. 3"},{"comment":"The definition of D as a log-likelihood ratio is clear, but the paper should state whether the p-values in Table 1 are obtained from Wilks' theorem, Monte Carlo simulations, or another prescription, since two entries are exactly 1.000.","section":"Sec. 4, Eq. (4.1)"},{"comment":"The caption 'the lower point (green) shows the lower limit at 95% CL' is ambiguous; please specify which curve is the likelihood and what the green point marks.","section":"Sec. 4, Fig. 3(b)"},{"comment":"The conclusion 'found 95% CL evidence for 100 TeV photons from five sources' should be rephrased as '95% CL lower limits on Ec above 100 TeV' to match the actual statistic and avoid the implication of direct event detections.","section":"Sec. 5"},{"comment":"Reference [7] (Bartoli et al.) is missing volume and page information, and the reference list would benefit from uniform formatting of arXiv identifiers.","section":"References"},{"comment":"The caption of Fig. 1(b) states that the colored bands give the systematic error of the HAWC spectra, but the text should clarify explicitly that these bands are taken from Ref. [1] and are not applied to the new high-energy limits in Table 1.","section":"Sec. 2"}],"recommendation":"major_revision","confidential_remarks":"This is a conference proceedings contribution, and the journal should consider whether the deferred systematic treatment and absence of event-level information meet its standards for archival publication. The paper overlaps with Ref. [1] and companion ICRC papers; the editor may wish to check for duplicate publication. The statistical methodology itself is sound and the results are of interest to the VHE gamma-ray community."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague — here's the quick read on 1908.11408.\n\nWhat's actually new: this is a conference proceedings in which HAWC applies its neural-network energy estimator to the Crab plus six other Galactic sources and reports the first lower limits on the cutoff energy Ec for those sources. Table 1 is the meat: at 95% CL, 2HWC J1825-134 has Ec > 253 TeV and 2HWC J1908+063 > 213 TeV; the Crab sits at 152 TeV. If those hold, HAWC has the strongest current evidence that at least two Galactic accelerators emit gamma rays past 200 TeV. The Crab spectrum itself extends beyond 100 TeV and agrees with IACT measurements within about 20% in the overlap region. That is a genuine, useful result.\n\nThe paper also does some things right. The profile-likelihood test for whether sources prefer a cutoff is honest: the p-values in Table 1 show no source prefers a finite Ec, so they are not selling a detection. The hard-cutoff model is admittedly non-physical, but they use it precisely because it is sensitive to excess counts above Ec without needing to know the spectral shape there. The forward-folding procedure is standard and described clearly enough to reproduce.\n\nThe soft spots are real but not fatal. The big one is systematics. Section 4 has a footnote deferring systematic errors to a future publication, and Table 1's limits are statistical only. The neural-network energy scale comes from simulations; if the absolute energy scale is off by 15%, the Ec limits move by a comparable fraction. For J1825-134, a downward shift still leaves the limit above 200 TeV, so the two-source headline might survive a few sigma of scale error—but we don't know because the size of that scale uncertainty isn't given. The stress-test note is right that this is the load-bearing caveat.\n\nThe abstract overstates things: 'observed photons above 200 TeV at 95% confidence' sounds like a direct photon count. The body correctly describes it as a lower limit on a cutoff energy derived from a model-dependent fit. The authors are not hiding this—it's all there in Section 4—but the abstract invites misreading.\n\nThe concern about circularity—using the fitted spectrum below Ec to predict mis-reconstructed events above Ec—is minor. They bin in tophats and the test is one-sided; it doesn't undermine the limits.\n\nWho this is for: gamma-ray astronomers and people working on Galactic cosmic-ray accelerators. It's a proceedings paper, so it isn't a full journal treatment; but the measurement is important enough that a referee would want to see the systematics folded in before taking the 200 TeV statement as definitive. If this crossed my desk as a journal submission I'd send it to review, and the main request would be a systematic error budget and a rewording of the abstract.","headline":"HAWC's first cutoff limits put two Galactic sources above 200 TeV, but the abstract overstates and systematic errors are deferred.","tokens_in":5800,"tokens_out":2907,"would_cite":true,"duration_ms":29530,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["95.85.Pw","98.70.Rz","95.55.Ka"],"model":"deepseek-v4-flash","headline":"HAWC reports photons above 200 TeV from galactic sources, with a 95% confidence lower limit on spectral cutoffs.","keywords":["HAWC","Crab Nebula","gamma-ray spectrum","very high energy photons","neural network energy reconstruction","spectral cutoff","Galactic sources","TeV astronomy"],"falsifier":"Suppose the actual energy scale of the NN estimator is shifted 20% lower (i.e., true photon energies are 20% higher than reconstructed for the same charge pattern). Then recomputing the profile likelihood with that corrected scale would lower each Ec by ~20%, moving the Crab below 100 TeV and possibly the 95% detection of >200 TeV into doubt. A direct test would be a calibration comparison with an IACT such as H.E.S.S. or MAGIC over 10-100 TeV, or the observation of an inverse-Compton bump of known spectral shape in an archival source.","tokens_in":4781,"feed_emoji":"🔭","tokens_out":1483,"duration_ms":14996,"temperature":0.7,"pith_summary":"This paper presents the first HAWC measurements of the highest-energy gamma-ray photons from the Crab Nebula and six other Galactic plane sources, using new neural-network energy reconstruction. The central claim is that HAWC has observed photons above 200 TeV at 95% confidence, setting lower limits on the energy cutoff for each source. If correct, this demonstrates that at least some Galactic particle accelerators produce gamma rays well beyond 100 TeV, constraining models of cosmic-ray acceleration and the Crab's emission mechanisms.","feed_headline":"HAWC detects photons above 200 TeV from Galactic sources","feed_subtitle":"New neural-network energy reconstruction pushes Crab spectrum beyond 100 TeV, setting cutoff limits.","key_machinery":"The central object is the hard-cutoff spectral test: a profile likelihood comparing a fit with no cutoff to a fit with a hard cutoff at energy Ec, designed to be sensitive to any excess above Ec without assuming the spectral shape above Ec. The analysis uses the neural-network (NN) energy estimator, which combines event size, zenith angle, shower location, and fractional charge in rings to achieve ~30% energy resolution above 10 TeV. The test separates true high-energy photons from lower-energy photon mis-reconstruction by forward-folding the fitted spectrum through the detector response.","core_discovery":"Using a new neural-network energy estimator calibrated on air-shower simulations, HAWC resolves photon energies from 1 TeV to beyond 100 TeV. For eight sources, the paper sets a lower limit on the cutoff energy Ec, above which emission is inconsistent with the background plus mis-reconstructed lower-energy events. At 95% confidence, the limits are 253 TeV for 2HWC J1825-134, 213 TeV for 2HWC J1908+063, and 152 TeV for the Crab; at the 3-sigma level, two sources still show emission above 100 TeV. The Crab spectrum agrees within 20% in flux with IACT measurements in the overlapping band, but extends the measured spectrum beyond 100 TeV.","pith_inferences":["The paper's hard-cutoff test is more model-independent than a full spectral fit, but the authors do not convert it into an absolute upper limit on photon energy; the 95% confidence claim of 'photons above 200 TeV' likely depends on the specific source with the highest Ec, and could be further strengthened by stacking sources or using a different energy estimator.","The neural-network energy estimator, if unbiased, could be extended to search for spectral cutoffs in fainter sources or to improve constraints on Lorentz-invariance violation, a connection the authors note but do not pursue here.","A testable extension is to apply the same hard-cutoff analysis to the upcoming HAWC data with more exposure; if the 200 TeV excess persists, the 95% lower limit on Ec should increase, while a large shift in the limits would indicate systematic bias in the energy scale."],"forward_implications":["If the 200 TeV photons are real, at least one Galactic source (likely 2HWC J1825-134 or 2HWC J1908+063) accelerates particles to energies above 200 TeV, challenging models that place a sharp cutoff below this energy.","The Crab Nebula spectrum extending beyond 100 TeV, rather than cutting off, favors emission models with hard synchrotron or inverse-Compton components that persist to extreme energies.","The lower limits on Ec, particularly the 253 TeV limit for 2HWC J1825-134, provide direct constraints on the maximum energy of accelerated electrons or ions in these systems, informing cosmic-ray acceleration theory.","The agreement of HAWC's Crab spectrum with IACT measurements within 20% validates the new NN energy reconstruction, supporting its use for future spectral analyses of other sources."],"supporting_citations":[{"why":"Provides the new energy estimators, the simulation-based calibration, and the Crab spectrum to above 100 TeV that this paper extends.","marker":"[1]"},{"why":"The previous HAWC energy estimation technique that the new estimators improve upon, used as a baseline for the Crab spectrum.","marker":"[2]"},{"why":"HEGRA's simple power-law fit to the Crab spectrum that the authors compare against their curved spectrum.","marker":"[8]"},{"why":"The Meyer inverse-Compton model of the Crab spectrum that the authors use as a reference for all other fits.","marker":"[9]"},{"why":"A more recent Fraschetti and Pohl IC model that provides an alternative theoretical prediction to compare with the HAWC data.","marker":"[11]"},{"why":"Thesis that details the hard-cutoff method for Lorentz invariance violation searches, repurposed here for spectral cutoffs.","marker":"[15]"},{"why":"HAWC's Lorentz invariance violation constraint using the same hard-cutoff technique, providing methodological context.","marker":"[16]"}],"fun_headline_variants":["HAWC neural net sees Crab photons above 200 TeV","Crab Nebula's spectrum extends past 100 TeV via HAWC","Machine learning reveals >200 TeV photons from Crab","HAWC's AI reconstruction pushes gamma-ray energy limit","Crab emits >200 TeV, HAWC sets cutoff constraints"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The claims depend entirely on the neural-network energy estimator correctly mapping measured PMT charges to true photon energies; if the energy scale is biased by even 15%, the quoted cutoff limits would shift correspondingly and could invalidate the 200 TeV detection.","fun_headline_variants_meta":{"raw":{"variants":["HAWC neural net sees Crab photons above 200 TeV","Crab Nebula's spectrum extends past 100 TeV via HAWC","Machine learning reveals >200 TeV photons from Crab","HAWC's AI reconstruction pushes gamma-ray energy limit","Crab emits >200 TeV, HAWC sets cutoff constraints"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000197,"raw_usage":{"total_tokens":1313,"prompt_tokens":841,"completion_tokens":472,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":457,"completion_tokens_details":{"reasoning_tokens":386}},"tokens_in":457,"tokens_out":472,"duration_ms":4534,"temperature":1.0,"reasoning_tokens":386,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T10:15:26.062901+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Suppose the actual energy scale of the NN estimator is shifted 20% lower (i.e., true photon energies are 20% higher than reconstructed for the same charge pattern). Then recomputing the profile likelihood with that corrected scale would lower each Ec by ~20%, moving the Crab below 100 TeV and possibly the 95% detection of >200 TeV into doubt. A direct test would be a calibration comparison with an IACT such as H.E.S.S. or MAGIC over 10-100 TeV, or the observation of an inverse-Compton bump of known spectral shape in an archival source.","supporting_citations":[{"cited_title":"2017 The Astrophysical Journal 843 39 URL http://iopscience.iop.org/article/10.3847/1538-4357/aa7555/meta","cited_arxiv_id":null,"evidence_quote":"The previous HAWC energy estimation technique that the new estimators improve upon, used as a baseline for the Crab spectrum."},{"cited_title":"2004 Astrophys","cited_arxiv_id":null,"evidence_quote":"HEGRA's simple power-law fit to the Crab spectrum that the authors compare against their curved spectrum."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The Meyer inverse-Compton model of the Crab spectrum that the authors use as a reference for all other fits."},{"cited_title":"Particle acceleration model for the broadband baseline spectrum of the Crab nebula","cited_arxiv_id":"1702.00816","evidence_quote":"A more recent Fraschetti and Pohl IC model that provides an alternative theoretical prediction to compare with the HAWC data."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Thesis that details the hard-cutoff method for Lorentz invariance violation searches, repurposed here for spectral cutoffs."},{"cited_title":"Constraints on Lorentz invariance violation using HAWC observations above 100 TeV","cited_arxiv_id":"1908.09614","evidence_quote":"HAWC's Lorentz invariance violation constraint using the same hard-cutoff technique, providing methodological context."}],"review_version":1}