{"id":"7639b09d-4ecb-4adf-aa6c-2ca03fb55081","arxiv_id":"2506.18500","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"low","formal_verification":"none","parameter_count":6,"one_line_summary":"For the TeV blazar PKS 2155-304, different X-ray data selections lead to very different modeled jet parameters, so claims about the blazar sequence must handle X-ray variability carefully.","lead":"This paper is a first step toward a TeV-scale version of the blazar sequence. It models one high-energy-peaked blazar, PKS 2155-304, with a standard emission model and finds that the inferred jet parameters change substantially depending on how the X-ray data are selected.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claimed parameter shift in Table 2 may reflect SSC degeneracy rather than data selection: δ is explicitly stated as poorly constrained, and log B changes by >3 dex without any reported uncertainties.","rationale":"The reader's verdict is CONDITIONAL, and my analysis does not change that: the paper remains a preliminary progress note whose central quantitative claim needs stronger support. However, my stress-test identifies a slightly different load-bearing condition than the reader's weakest assumption. The reader focused on whether the assembled SED truly represents an 'average activity state'; I focus on whether the differences in Table 2 are real or are dominated by model degeneracy. The two are related—both make the fitted parameters non-unique—but the degeneracy concern is more directly fatal because it applies even if the data selections are both perfectly representative. The manuscript itself flags that δ is poorly constrained, and no uncertainties are reported for any parameter. The log B difference of 3.2 dex is the most extreme parameter shift and is not addressed by the authors' caveat, so it is the natural place to probe. The proposed concrete test would settle the issue by producing posterior distributions or by demonstrating whether a common δ can fit both data selections. If the test shows overlapping credible regions, the claim that data selection changes the physical jet properties is unsupported; if the regions are disjoint, the claim is validated. Given the preliminary nature of the contribution, the CONDITIONAL verdict remains appropriate, with the condition being that the full study supplies parameter uncertainties or otherwise demonstrates the significance of the reported shifts. I therefore set verdict_should_be to UNCHANGED, meaning I do not move the reader's verdict.","tokens_in":962,"tokens_out":899,"duration_ms":60201,"concrete_test":"Run a full MCMC or dense grid scan over the MMDC/SSC parameter space for both X-ray data selections of PKS 2155-304, and compute marginalized posterior distributions and joint credible regions for δ, log B, and t_var. If the 68% (or 95%) contours for the two selections overlap substantially—especially for log B, whose point estimates differ by 3.2 dex—then the reported parameter shift is not statistically significant and the central claim fails. As a cross-check, fix δ to a grid of values spanning 5–100 and refit the remaining parameters for both data selections; if comparable goodness-of-fit is achieved with a single δ for both selections, the claimed δ difference is a degeneracy artifact rather than an effect of data selection.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing concern is that the paper's central claim—that X-ray data selection materially changes fitted SSC parameters for PKS 2155-304—may be an artifact of the well-known degeneracies in one-zone SSC models rather than a physical response to the data selection. Table 2 reports point estimates only, with no error bars or credible intervals. The authors themselves state in Sect. 5 that 'the large difference in the resulting best parameters can be ascribed to its poorly constrained distribution in the modeling (see Fig. 4b), which makes these results less reliable' for δ. The factor-of-two change in δ (21.47 to 43.06) is therefore not robustly attributable to the X-ray selection without a posterior comparison. The change in log B is even larger—from 1.70 to -1.51, a 3.2 dex shift—yet the corner plot shown does not demonstrate that log B is independently constrained. In one-zone SSC models, B, δ, and source size are strongly degenerate; the two 'best-fit' points may lie along the same degenerate ridge. If so, the conclusion that the data selections correspond to different physical jet properties is not supported, and the broader implication for the TeV blazar sequence loses its evidentiary basis. This concern is directly flagged by the manuscript's own admission about δ, which strengthens its relevance.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This proceedings paper proposes a new way to revisit the blazar sequence by binning TeV-detected BL Lacs into one-decade synchrotron-peak-frequency bins and selecting one representative source per bin, then modeling each source with a one-zone synchrotron self-Compton (SSC) model. The paper focuses on PKS 2155-304, the representative source for the bin 10^15 < nu_syn < 10^16 Hz, and compares SSC fits obtained with two different X-ray data selections: an average of archival X-ray data versus a single X-ray observation. The resulting best-fit parameters (Table 2) differ substantially, most notably in the Doppler factor delta (21.47 versus 43.06) and the magnetic field log B (1.70 versus -1.51), and the authors interpret this as evidence that data selection materially affects the modeled jet properties, with implications for the physical interpretation of the TeV blazar sequence. The paper explicitly acknowledges that delta is poorly constrained and that the source is highly variable, and it frames the work as a preliminary step toward modeling a larger sample.","tokens_in":5370,"tokens_out":3870,"duration_ms":45882,"significance":"If validated, the proposed frequency-binned selection protocol would provide a systematic framework for modeling TeV BL Lacs in the context of the blazar sequence, and the emphasis on average-state data would address a known weakness of archival SED modeling. The paper's strengths include its transparent presentation of the source-selection criteria, its use of publicly available databases (4LAC-DR2, MMDC, STeVECat), and its explicit caveats about X-ray variability and the poor constraint on delta. The manuscript also honestly reports the tension between two reasonable data selections rather than hiding it. However, the central quantitative claim--that the X-ray selection changes the physical jet parameters--currently rests on point estimates without uncertainty quantification, and the paper itself flags the degeneracy-related unreliability of delta. Because the paper aims at a physical interpretation of the blazar sequence, the absence of credible intervals or a model-comparison test limits the evidentiary weight of the reported parameter shifts.","major_comments":[{"comment":"The central claim that the two X-ray selections yield materially different SSC parameters is not robustly established because Table 2 reports only point estimates without uncertainties or credible intervals. The text itself states that the difference in delta 'can be ascribed to its poorly constrained distribution in the modeling (see Fig. 4b), which makes these results less reliable,' and the change in log B from 1.70 to -1.51 (3.2 dex) is presented without demonstrating that log B is independently constrained. Since one-zone SSC models are known to have strong degeneracies among B, delta, and source size, the two best-fit points could lie on the same degenerate ridge; a quantitative comparison of the posterior distributions (e.g., credible intervals, profile likelihoods) or fits with delta fixed to a grid of values is needed to attribute the parameter shifts to the data selection rather than to model degeneracy.","section":"Sect. 5 and Table 2"},{"comment":"The claim that the assembled non-simultaneous SEDs represent an 'average activity state' is not demonstrated quantitatively. The paper acknowledges that simultaneous multi-wavelength data were not always available and that X-ray observations were gathered from different periods, but no criterion (e.g., X-ray flux relative to the long-term distribution, or an activity-state indicator) is given for either the averaged selection or the selected single observations. Given the high variability of PKS 2155-304 shown in Fig. 3, if neither selection actually corresponds to the average state, the fitted jet parameters and the implication for the TeV blazar sequence are not physically meaningful. The authors should either validate the selected states against the source's long-term X-ray light curve or frame the results explicitly as bracketing states rather than average-state properties.","section":"Sect. 3"},{"comment":"The modeling tool MMDC is trained by the same group (including a co-author) on SOPRANO simulations, and no independent cross-check against other SSC fitting codes or Monte Carlo methods is presented. Because the paper aims at a physical interpretation, the absolute parameter values in Table 2 inherit any systematic bias of the training set; a validation run with an independent code (e.g., a standard Markov chain Monte Carlo implementation) on at least PKS 2155-304 would demonstrate that the reported differences are not artifacts of the specific CNN architecture or training data.","section":"Sect. 4"}],"minor_comments":[{"comment":"The in-text references to table numbers are inconsistent: Sect. 4 refers to 'Table 4' for the modeling results while the displayed table is labeled Table 2, and Sect. 2 refers to 'Table 2' for representative sources while the displayed table is labeled Table 1.","section":"General"},{"comment":"There is a typo in the sentence 'VHEγ-ray data were taken from from the Spectral TeV Extragalactic Catalog'--the word 'from' is duplicated.","section":"Sect. 3"},{"comment":"The word 'Netowrks' in the description of MMDC is misspelled and should be 'Networks'.","section":"Sect. 4"},{"comment":"The phrase 'continuous of sources in X-rays' is unclear and should be reworded, for example as 'continuous monitoring of sources in X-rays'.","section":"Sect. 5"},{"comment":"The corner plot in Fig. 4 would be easier to interpret if the parameter ranges and contours were labeled with units (e.g., delta is dimensionless, B in Gauss), and if the posterior distributions were overlaid for both X-ray selections.","section":"Fig. 4"}],"recommendation":"major_revision","confidential_remarks":"This is a preliminary proceedings contribution, and the requested uncertainty quantification may be more than is typical for this venue. However, the central claim about data-selection-driven parameter shifts needs at least a posterior comparison or an explicit reframing as a bracketing exercise; otherwise the conclusions outrun the evidence presented. A minor revision could be sufficient if the authors clearly downgrade the claim to a demonstration of sensitivity rather than a determination of physical jet properties."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This is a short proceedings paper, and it reads like one: a clear description of a planned program, one worked example, and a heavily caveated result. The frequency-binning of the blazar sequence is a sensible alternative to luminosity binning, and the paper is refreshingly honest that the two X-ray selections give quite different best-fit parameters—and that the Doppler factor, in particular, is poorly constrained.\n\nWhat is genuinely new: binning by synchrotron peak frequency rather than luminosity, and choosing representative sources rather than averaging SEDs. That is a reasonable methodological twist and worth keeping in mind. The comparison of averaged versus single-observation X-ray data for PKS 2155-304 is also a useful caution for anyone building SEDs from archival data, and the corner plot lets the reader see the degeneracies directly.\n\nThe soft spots are real, though. The paper's central claim—that data selection materially changes the fitted jet parameters—is not supported by the numbers. Table 2 gives point estimates with no error bars. The Doppler factor shifts from 21.47 to 43.06, but the authors themselves say it is poorly constrained and the results are less reliable. The magnetic field swings from log B = 1.70 to -1.51, a 3.2 dex jump, which is exactly what you expect if the fit is just sliding along the degenerate B–δ ridge in a one-zone SSC model. So the two ``best-fit'' points may correspond to the same physical solution, not two distinct states. The stress-test note is right about this, and the manuscript's own admission about δ strengthens the point.\n\nThere is also a mismatch between the title and the content: no TeV blazar sequence is delivered here, just one source and a plan. That is fine for a proceedings, but the framing oversells. The MMDC circularity is a minor issue—the tool is published, but it was built by the same group, so independent cross-checks with a traditional fitting code would add confidence. The use of non-simultaneous archival data for a highly variable source also makes the ``average state'' concept fragile, as the authors concede.\n\nBottom line: as a progress note, this is acceptable and should not be desk-rejected. But it should be treated as a description of work in progress, not as evidence that X-ray selection changes the physics. The referee should ask for credible intervals or an explicit statement that the two selections cannot be distinguished given the degeneracies. A light referee is enough; a heavy one is overkill.","headline":"Honest progress note on a frequency-binned TeV blazar sequence, but the one quantitative result is undermined by SSC degeneracy the authors themselves flag.","tokens_in":5906,"tokens_out":1898,"would_cite":false,"duration_ms":23200,"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":"For PKS 2155-304, choosing which X-ray data count as 'average' doubles the fitted Doppler factor and flips the magnetic field by three orders of magnitude.","keywords":["blazars","BL Lac objects","blazar sequence","Synchrotron Self Compton","PKS 2155-304","TeV blazars","spectral energy distribution","X-ray variability"],"falsifier":"Run both SED selections through the full SOPRANO simulation code rather than the convolutional-neural-network emulator and re-fit; if the best-fit Doppler factor and magnetic field no longer show the separation in Table 2, the claimed data-selection sensitivity would be an artifact of the surrogate model. In parallel, an SED built from strictly simultaneous X-ray and very-high-energy observations of PKS 2155-304 taken during a single quiescent epoch would test which archival selection, if either, approximates the source's true average state.","tokens_in":4831,"feed_emoji":"🔭","tokens_out":8998,"duration_ms":95337,"temperature":0.7,"pith_summary":"This paper is an early step toward explaining the blazar sequence—the anticorrelation between blazar luminosity and the frequency of their spectral peaks—in physical terms, by modeling TeV-detected BL Lac objects one by one in a synchrotron self-Compton (SSC) framework. For the representative high-energy-peaked source PKS 2155-304, the authors show that the fitted SSC parameters depend strongly on how the archival X-ray data are selected: averaging the X-ray data versus using single observations in an average state yields a Doppler factor that doubles, from 21.47 to 43.06, and a magnetic field that changes by roughly three orders of magnitude. Since both selections are reasonable and the source is highly variable in X-rays, the message is that an 'average-state' SED of a TeV blazar is not well-defined from archival data, and this ambiguity must be resolved before any physical interpretation of the TeV blazar sequence can be trusted. The wider goal is to model the remaining representative sources in five synchrotron-peak bins and look for trends in the fitted jet parameters.","feed_headline":"Data selection doubles a blazar's fitted Doppler factor","feed_subtitle":"Averaging vs single X-ray observations gives SSC jet parameters that differ enough to complicate the TeV blazar sequence.","key_machinery":"The machinery is the one-zone Synchrotron Self-Compton (SSC) model, in which a single relativistic blob of electrons produces both the low-energy synchrotron bump and the high-energy bump by up-scattering those same synchrotron photons through inverse Compton. The model is applied through an online fitting tool whose convolutional neural network, trained on spectra simulated with the SOPRANO code, returns best-fit parameters and corner plots over the free parameters (electron luminosity, injection index, Doppler factor, magnetic field, blob size, variability timescale). The contrast between two X-ray data selections for PKS 2155-304 is what carries the argument: the same model, the same source, and two defensible data choices produce widely separated parameter values.","core_discovery":"The paper's central claim, stated on its own terms, is that data selection materially affects the modeled jet properties of PKS 2155-304 within a simple SSC model. Using the MMDC fitting tool, the authors fit the same source twice: once with averaged X-ray data and once with single X-ray observations corresponding to an average activity state. The resulting best-fit parameters are substantially different—most notably the Doppler factor (δ = 21.47 vs 43.06) and the magnetic field (log(B/G) = 1.70 vs −1.51)—while the electron luminosity and energy-break parameters stay close. The authors attribute the large Doppler-factor shift to a poorly constrained parameter distribution, and note that it is not trivial to say which selection is correct. Thus the paper establishes a caution: any physical reading of the blazar sequence from SSC fits of archival SEDs is sensitive to the choice of which observations represent the source.","pith_inferences":["If the sensitivity seen in PKS 2155-304 holds for other representatives, then the trend that a future 'TeV blazar sequence' reveals could be partly an artifact of how each source's archival X-ray data are selected; binning by synchrotron peak frequency does not by itself control for this.","A natural stress test would be to fit multiple alternative realizations of the same source—say X-ray data split by flux state—and check whether the inferred SSC parameters trace the activity state continuously; if they do, the fits could be inverted to estimate the state of the source at the time of each observation.","The large swing in magnetic field from log(B/G) = 1.70 to −1.51, if real, has consequences beyond the sequence: it changes estimates of the jet's particle energy budget and of the γ-ray emission region's location, so the data-selection ambiguity could affect TeV BL Lac population studies generally."],"forward_implications":["The planned fits of representative sources in the other synchrotron-peak bins will only support a physical interpretation of the blazar sequence if the fitted parameters are stable against equivalent data-selection choices; the PKS 2155-304 result puts the onus on showing that stability.","Reported SSC parameters for PKS 2155-304 should carry a note about the X-ray selection: the Doppler factor and magnetic field cannot be quoted as single numbers without specifying whether X-ray data were averaged or taken from single observations.","Because the Doppler factor distribution is poorly constrained, conclusions that use the fitted δ for this source—for example, estimates of jet beaming or jet power—should be treated as provisional.","Future continuous X-ray monitoring of TeV BL Lacs, or tools that filter observations by epoch, would replace the two-selection ambiguity with a well-defined average state."],"supporting_citations":[{"why":"Establishes the blazar sequence anticorrelation that this work aims to reinterpret physically.","marker":"Fossati et al. (1998)"},{"why":"Confirms the sequence with Fermi-LAT sources and provides the luminosity-bin method the authors replace with synchrotron-peak bins.","marker":"Ghisellini et al. (2017)"},{"why":"Introduces the one-zone SSC model used for the SED fits.","marker":"Maraschi et al. (1992)"},{"why":"Supplies the MMDC database and modeling platform used to build SEDs and run the SSC fits.","marker":"Sahakyan et al. (2024)"},{"why":"Describes the convolutional neural network underlying the modeling tool's fast parameter scan.","marker":"Bégué et al. (2024)"},{"why":"Provides the SOPRANO code whose simulated blazar spectra train the neural network.","marker":"Gasparyan et al. (2022)"},{"why":"Previous extension of the sequence analysis to TeV blazars that motivates the present approach.","marker":"Prandini & Ghisellini (2022)"},{"why":"STeVECat catalog from which the very-high-energy gamma-ray data for the SEDs are taken.","marker":"Gréaux et al. (2023)"}],"fun_headline_variants":["Data choice flips Doppler factor in blazar SED fits","X-ray averaging alters blazar jet parameters twofold","Blazar sequence muddied by data selection effects","PKS 2155-304 shows data-driven parameter shift","SSC fits sensitive to X-ray data averaging"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing assumption is that the non-simultaneous archival data selected for PKS 2155-304, under either X-ray treatment, truly represents the source in an average activity state; if neither selection does, the fitted jet parameters are not physical.","fun_headline_variants_meta":{"raw":{"variants":["Data choice flips Doppler factor in blazar SED fits","X-ray averaging alters blazar jet parameters twofold","Blazar sequence muddied by data selection effects","PKS 2155-304 shows data-driven parameter shift","SSC fits sensitive to X-ray data averaging"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000121,"raw_usage":{"total_tokens":1077,"prompt_tokens":912,"completion_tokens":165,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":528,"completion_tokens_details":{"reasoning_tokens":88}},"tokens_in":528,"tokens_out":165,"duration_ms":2653,"temperature":1.0,"reasoning_tokens":88,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T23:16:19.608768+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run both SED selections through the full SOPRANO simulation code rather than the convolutional-neural-network emulator and re-fit; if the best-fit Doppler factor and magnetic field no longer show the separation in Table 2, the claimed data-selection sensitivity would be an artifact of the surrogate model. In parallel, an SED built from strictly simultaneous X-ray and very-high-energy observations of PKS 2155-304 taken during a single quiescent epoch would test which archival selection, if either, approximates the source's true average state.","supporting_citations":[{"cited_title":"1998, Monthly Notices of the Royal Astronomical Society, 299, 433","cited_arxiv_id":null,"evidence_quote":"Establishes the blazar sequence anticorrelation that this work aims to reinterpret physically."},{"cited_title":"2017, , 469, 255","cited_arxiv_id":null,"evidence_quote":"Confirms the sequence with Fermi-LAT sources and provides the luminosity-bin method the authors replace with synchrotron-peak bins."},{"cited_title":"1992, , 397, L5","cited_arxiv_id":null,"evidence_quote":"Introduces the one-zone SSC model used for the SED fits."},{"cited_title":"2024, Astron","cited_arxiv_id":null,"evidence_quote":"Supplies the MMDC database and modeling platform used to build SEDs and run the SSC fits."},{"cited_title":"2022, Mon","cited_arxiv_id":null,"evidence_quote":"Provides the SOPRANO code whose simulated blazar spectra train the neural network."}],"review_version":1}