{"id":"b68d9e45-3e2f-4f93-b413-1226ccaaea69","arxiv_id":"2506.08497","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"A new Fermi-LAT catalog of 275 very-high-energy active galactic nuclei at high Galactic latitudes, including 204 new detections, and a luminosity function for BL Lac blazars.","lead":"This paper uses 16 years of Fermi satellite data to find 275 active galaxies that emit very high energy gamma rays above 100 GeV, more than doubling the known high-energy extragalactic source list. The resulting catalog gives astronomers a target list for next-generation gamma-ray observatories and a new measurement of how common these extreme galaxies are.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The BL Lac luminosity function is presented without stating how sources missing redshifts are treated or how completeness is corrected, so the break at L*=1e44 erg/s and the 6.5e-7 Mpc^-3 density are not yet substantiated.","rationale":"The reader's weakest assumption is the 4FGL parent-sample selection, which is a real caveat: the survey is not a blind all-sky search, so any VHE AGN absent from 4FGL would be missed. However, in practice this is unlikely to be numerically dominant, because an AGN that produces multiple >100 GeV photons over 16 years would usually produce many more photons below 100 GeV for any plausible blazar spectral index, making a 4FGL detection very likely. The more pressing issue is the luminosity function, which is the headline quantitative result beyond the catalog itself. The paper never states how sources without measured redshifts are handled, how the flux-limited nature of the sample is corrected, or whether density evolution is included. A substantial fraction of the new BL Lacs have no redshift in Tables 1-4, so the LF is built on a biased subset. This concern is concrete and testable with the published tables. The catalog detection claim itself appears sound, so the verdict should remain CONDITIONAL pending a documented LF analysis, rather than being accepted or rejected outright.","tokens_in":29621,"tokens_out":9648,"duration_ms":123301,"concrete_test":"Recompute the BL Lac LF from the published tables using a 1/Vmax estimator with three treatments: (a) only sources with known redshifts, (b) a survival-analysis treatment of sources without redshifts (treated as lower limits), and (c) the paper's stated 90% completeness flux limit (1.3e-12 erg/cm2/s) applied consistently. Fit a broken power law in each case and compare the break luminosity and normalization. If the break moves by more than 0.3 dex or the source density changes by more than 50% between treatments (a) and (b)/(c), the claimed LF is not robust.","verdict_should_be":"UNCHANGED","load_bearing_attack":"In Section 3.2, the paper claims that the BL Lac luminosity function is 'well measured' and shows a broken power law with a break near L*=1e44 erg/s and a source density (6.5±0.5)e-7 Mpc^-3. However, no LF estimator, completeness correction, or redshift-incompleteness treatment is described. Inspection of Tables 1-4 shows that many newly detected BL Lacs have no redshift (z='-1' in the tables), and those sources cannot contribute to luminosity-volume bins in Fig. 4. If they are simply omitted, the LF normalization is biased low and the break position becomes sensitive to the flux limit, because the missing-redshift sources are preferentially fainter or more distant. The text's visual argument that the deficit at L<L* is not due to sensitivity (based on Fig. 3) applies only to the subset with known redshifts. No density evolution or EBL-attenuation correction is described for the LF, so the break could reflect the survey horizon rather than a genuine feature of the BL Lac population. This directly undermines the quantitative population claim, even if the catalog itself is reliable.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a catalog of very-high-energy (VHE, E>100 GeV) AGN at Galactic latitudes |b|>10 deg, constructed by searching for clusters of Fermi-LAT photons around positions of 4FGL sources. It reports 175 sources detected at >4 sigma and 100 additional sources at >3 sigma, with a claimed 90% completeness at a flux of 1.3e-12 erg/cm2/s. The catalog is used to derive a BL Lac luminosity function described as a broken power law with break at L* ~ 1e44 erg/s and normalization (6.5+/-0.5)e-7 Mpc^-3. The paper also discusses subsamples of radio galaxies, FSRQs, extreme blazars, high-redshift sources, and unidentified objects, and suggests applications for CTAO planning and EBL/IGMF studies.","tokens_in":29798,"tokens_out":8924,"duration_ms":91924,"significance":"If the catalog is reliable, it is a valuable resource: it provides a homogeneous, well-defined VHE AGN sample from 16 years of Fermi-LAT data, with explicit false-positive estimates, per-source spectra, and detailed discussion of high-redshift sources. It would be the first systematic VHE AGN catalog from a space-based instrument and could guide CTAO observations and EBL/IGMF studies. The luminosity-function claim, if substantiated, would be an important population measurement. However, the LF analysis is currently underdocumented and the 'unbiased survey' claim is not fully supported, so the quantitative population results should be treated with caution until the analysis is described in detail.","major_comments":[{"comment":"The broken-power-law BL Lac luminosity function quoted in the abstract and in Sec. 3.2 is not substantiated. The paper gives no LF estimator, no completeness correction, and no treatment of the many sources with z=-1 in Tables 1-4. If missing-redshift sources are omitted from the luminosity-volume bins, the derived normalization (6.5+/-0.5)e-7 Mpc^-3 and the break at L*~1e44 erg/s are biased, because the omitted sources are preferentially fainter or more distant. The argument in the text that the low-luminosity deficit is not a sensitivity effect (based on Fig. 3) applies only to the redshift-known subset. Please provide a quantitative LF derivation, including the redshift-completeness correction, the EBL correction, and the assumed density evolution (or justify the no-evolution assumption).","section":"Sec. 3.2, Figs. 3 and 4"},{"comment":"The claim in Sec. 3.1 that the catalog provides a sky survey 'not biased by selection effects' is contradicted by the search procedure described in Sec. 2, which uses positions of sources from the 4FGL catalog as priors. The parent sample is GeV-selected, so VHE AGN too faint in the 0.1-100 GeV band to enter 4FGL are missed. This selection effect should be quantified (e.g., by an estimate of the fraction of VHE sources missing from 4FGL, or a comparison with a blind search) and propagated into the completeness and LF claims.","section":"Sec. 2 and Sec. 3.1"},{"comment":"The 90% completeness limit at F=1.3e-12 erg/cm2/s is derived from a single average photon count assuming E=100 GeV and simplified Monte Carlo. This does not account for the distribution of spectral indices or the energy dependence of the PSF and effective area. Because the detection thresholds use fixed 0.1 and 0.4 degree apertures, hard-spectrum sources (more photons at high energy, better PSF) are favored over soft-spectrum sources at equal flux. Please quantify the spectral-index dependence of the completeness and the resulting selection function.","section":"Sec. 3.1"}],"minor_comments":[{"comment":"Typo: 'probability to fine one photon' should read 'probability to find one photon'; also 'simulaitons' in Sec. 3.1 should be 'simulations'.","section":"Sec. 2 and Sec. 3.1"},{"comment":"The number of extreme blazars is given as 63 in the abstract and Sec. 3.3 (22 previously known + 41 new), but the Discussion states 49. Please reconcile.","section":"Sec. 3.3 and Discussion"},{"comment":"The tables use z=-1 for missing redshifts, but this is not explained in the table captions or text. Add a footnote describing the convention and stating the number of sources without redshifts.","section":"Tables 1-4"},{"comment":"Fig. 4 does not show error bars or the broken-power-law fit; please show the data with uncertainties and the fit curve, and report the fit parameters (kappa1, kappa2, L*, Phi*) with uncertainties.","section":"Sec. 3.2"},{"comment":"Typo: 'PSK 0454-234' should be 'PKS 0454-234'.","section":"Sec. 3.5"},{"comment":"In the paragraph on 4FGL J1955.3-5032, the source name is written 'FGL J1955.3-5032'; the prefix should be '4FGL'.","section":"Sec. 3.6"}],"recommendation":"major_revision","confidential_remarks":"The catalog construction and the per-source data are valuable and likely citable, but the luminosity-function section is not yet at the standard required for the quantitative claims made in the abstract. The 'unbiased survey' statement needs to be tempered or supported with a quantitative selection-function analysis. The internal inconsistency in extreme-blazar counts (63 vs 49) suggests the manuscript needs careful proofreading. The paper is within scope for A&A and the authors are well known in the field; I recommend major revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this is a useful catalog paper. The 275-source VHE AGN list with 204 new detections is a real step up from the earlier 2011/2015 Fermi surveys and from TeVCat, and it gives the field a reference sample for CTAO planning and for extreme blazar follow-up. The detection statistics are transparent: the photon-clustering criteria are spelled out, the false-positive budget is estimated (about 3 in the 3-sigma tables), and the completeness estimate (90% at 1.3e-12 erg/cm2s) is reasonable, even if it only uses a single effective photon energy. The high-redshift section is also honest — the authors flag that the apparent spectral hardenings could be wrong redshifts or EBL-model limitations, not a detection artifact.\n\nThe soft spot is the luminosity function. Section 3.2 says the BL Lac LF is 'well measured' and gives a break at L*~1e44 erg/s and a spatial density of 6.5e-7 Mpc^-3, but the paper never describes a proper LF estimator, completeness correction, or how sources without redshifts are treated. Looking at the tables, a large fraction of the new BL Lacs have z=-1. If those are simply dropped from the luminosity-volume bins, the normalization is biased and the break position becomes sensitive to the flux limit. The paper's argument that the deficit below L* is not sensitivity-driven (based on Fig. 3) only holds for the subset with known redshifts. This is a significant gap because the LF is one of the headline results. It needs a real treatment — or at least a clear statement of assumptions and a caveat.\n\nThe other issue is the 'unbiased sky survey' claim. The search is performed only around 4FGL source positions. In practice most VHE AGN are bright enough in the GeV band to be in 4FGL, so this may not change the catalog much, but the wording is stronger than the method supports. That is a minor overstatement, not a fatal flaw.\n\nWho should read this: gamma-ray astronomers and anyone planning CTAO or IACT observations. Take the catalog, treat the LF as a preliminary result. This deserves peer review — the catalog is important enough that a referee should ask for a proper LF analysis rather than see the paper desk-rejected.","headline":"A substantial new VHE AGN catalog from 16 years of Fermi-LAT data, with an under-supported BL Lac luminosity function that needs more work before the population claims land.","tokens_in":30409,"tokens_out":2968,"would_cite":true,"duration_ms":33738,"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":"The paper claims Fermi-LAT's 16-year exposure, searched for clusters of >100 GeV photons around 4FGL sources, yields a 90%-complete catalog of 275 very-high-energy AGN at high Galactic latitudes and an associated BL Lac luminosity…","keywords":["very-high-energy gamma rays","active galactic nuclei","BL Lac objects","Fermi-LAT","gamma-ray survey","luminosity function","extragalactic background light","blazars"],"falsifier":"Perform the same 16-year Fermi-LAT search without anchoring to 4FGL positions, i.e., look for any clusters of E>100 GeV photons at |b|>10° with no catalog counterpart. A population of such orphan clusters above the claimed 1.3×$10^{-12}$ erg/$cm^{2}$ s completeness limit would show the catalog is not the unbiased census it claims to be.","tokens_in":29363,"feed_emoji":"🔭","tokens_out":7747,"duration_ms":81551,"temperature":0.7,"pith_summary":"The paper claims that Fermi-LAT's sixteen years of continuous sky scanning now acts as an unbiased survey for very-high-energy (VHE, >100 GeV) active galactic nuclei. By counting clusters of >100 GeV photons inside the point-spread function of sources from the 4FGL catalog, the authors identify 175 AGN at high significance and another 100 at the 3σ level in the high-latitude sky, 204 of which were not previously known as VHE emitters. They find the catalog is 90% complete at a flux of 1.3×$10^{-12}$ erg/$cm^{2}$ s and use the dominant BL Lac population to measure a luminosity function that is a broken power law with a break at L* ≈ $10^{44}$ erg/s and a source density of (6.5±0.5)×$10^{-7}$ $Mpc^{-3}$. If correct, this converts the VHE sky from a collection of pointed IACT discoveries into a statistically defined population, and provides a ready-made target list for the next generation of imaging atmospheric Cherenkov telescopes and for studies of the extragalactic background light.","feed_headline":"16-year Fermi survey catches 275 high-energy AGN","feed_subtitle":"Catalog is 90% complete, adds 204 new VHE blazars, and fixes the BL Lac luminosity function break.","key_machinery":"The load-bearing object is the photon-cluster statistic: counting E>100 GeV photons within the 68% and 95% PSF containment circles around each 4FGL source, with an analytic chance-coincidence probability p1 ≈ 0.0122 per photon inside 0.1°, so that three photons already give a ~4σ detection. The catalog's population statement then rests on the broken power-law luminosity function dN/dL ∝ $L^{{-κ}}$ of BL Lacs, whose break at L* ≈ $10^{44}$ erg/s separates the flux-limited part of the sample from the intrinsically rare bright end.","core_discovery":"Using the public Fermi-LAT data from September 2008 to March 2025, the paper searches for spatial clustering of photons with E>100 GeV around the positions of the 5071 4FGL sources at |b|>10°. A source is retained with high confidence when it has three or more photons inside the 68% PSF containment circle (chance probability ~1.8×$10^{-6}$), or two photons inside 68% plus one inside the 95% circle; weaker criteria select the 3σ sample. This yields 175 sources above 4σ and 100 above 3σ, of which only 71 were previously reported by IACTs. The authors argue the sample is an unbiased sky survey, 90% complete at 1.3×$10^{-12}$ erg/$cm^{2}$ s, and derive from the 233 BL Lacs a luminosity function dN/dL ∝ $L^{{-κ}}$ with a break from κ<2 to κ>2 at L* ≈ $10^{44}$ erg/s, corresponding to a spatial density (6.5±0.5)×$10^{-7}$ $Mpc^{-3}$. They also report seven sources at z>1 whose EBL-corrected spectra harden above 100 GeV, which they attribute either to wrong redshifts or to a factor-of-two overestimate of the high-redshift EBL opacity.","pith_inferences":["If the catalog is as complete as claimed, the historical IACT source list was strongly biased toward bright, nearby, hard-spectrum blazars; population models built on that anecdotal list would need to be redone with this survey-defined sample.","The 4FGL-anchored search could miss a hypothetical class of 'orphan' VHE emitters, such as BL Lacs with steep GeV spectra or one-off flares; a blind cluster search would test this directly.","The luminosity-function break at ~10^44 erg/s coincides with the typical divide between FR I and FR II radio galaxies; if some of the 20 unclassified AGN and blazar candidates are radio galaxies, the total VHE power of the radio-galaxy population, which the paper estimates may rival BL Lacs, could be constrained precisely.","If the high-z hardening is genuine, VHE spectra of z>1 blazars could serve as a redshift-independent probe of EBL evolution, and the three sources with disputed redshifts (GB6 J0045+2127, Ton 0396, RBS 1432) are the ones to target with optical spectroscopy."],"forward_implications":["The 275-source catalog gives current and next-generation IACT arrays a pre-vetted target list; a survey of all sources at about 50 hours each would cost roughly 1.4×10^4 hours, on the scale of a decade of allocated observation time.","The BL Lac luminosity function implies a total VHE power injection of (1.41±0.11)×10^37 erg/s/Mpc^3, which extrapolated over a Hubble time reproduces the measured extragalactic gamma-ray background at 100 GeV.","The 63 extreme blazars, 41 of them new, provide a much larger systematic sample for intergalactic magnetic field searches than the handful of sources previously studied in depth.","The seven z>1 detections show that Fermi-LAT can find VHE sources beyond the gamma-ray horizon probed by current IACTs, and their spectral hardening in EBL-corrected spectra can be removed by halving the EBL optical depth at z≈1.5–2, within the stated uncertainty of the model."],"supporting_citations":[{"why":"Supplies the 4FGL parent source list, positions, spectral indices, and source classifications that the VHE cluster search is anchored to.","marker":"Ballet et al. 2023"},{"why":"Provides the EBL model used for absorption correction of VHE fluxes and luminosities, and the reference for the claimed factor-of-two uncertainty at high redshift.","marker":"Saldana-Lopez et al. 2021"},{"why":"Provides the alternative EBL model that produces stronger spectral hardening for high-redshift sources and the z≈1 optical-depth benchmark.","marker":"Franceschini & Rodighiero 2017"},{"why":"Establishes the earlier all-sky Fermi-LAT search method and source list that this work extends with sixteen-year exposure.","marker":"Neronov et al. 2015"},{"why":"Provides the spectroscopic redshift limits used to assess the disputed redshifts of the high-z VHE sources.","marker":"Shaw et al. 2013"},{"why":"Supplies the cosmological parameters used to convert fluxes and redshifts into luminosities and comoving densities.","marker":"Planck Collaboration et al. 2020"},{"why":"Provides the high-synchrotron-peak (extreme blazar) classification used to tag the 'E' sources in the catalog.","marker":"Chang et al. 2019"}],"fun_headline_variants":["Fermi survey reveals 275 VHE AGN, 204 new","16-year Fermi data yields 275 VHE AGN, 90% complete","New VHE AGN catalog: 275 sources, 204 never seen before","Fermi finds 275 high-energy AGN, mostly BL Lacs","VHE AGN catalog from Fermi includes 275 sources"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"Load-bearing premise: every VHE gamma-ray source in the field is bright enough in the 0.1–100 GeV band to be listed in the 4FGL catalog, since the cluster search is performed only around those positions; a VHE AGN population too faint in GeV would be absent from the survey and from the luminosity function.","fun_headline_variants_meta":{"raw":{"variants":["Fermi survey reveals 275 VHE AGN, 204 new","16-year Fermi data yields 275 VHE AGN, 90% complete","New VHE AGN catalog: 275 sources, 204 never seen before","Fermi finds 275 high-energy AGN, mostly BL Lacs","VHE AGN catalog from Fermi includes 275 sources"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00072,"raw_usage":{"total_tokens":3380,"prompt_tokens":1241,"completion_tokens":2139,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":857,"completion_tokens_details":{"reasoning_tokens":2040}},"tokens_in":857,"tokens_out":2139,"duration_ms":18825,"temperature":1.0,"reasoning_tokens":2040,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T05:10:09.989108+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Perform the same 16-year Fermi-LAT search without anchoring to 4FGL positions, i.e., look for any clusters of E>100 GeV photons at |b|>10° with no catalog counterpart. A population of such orphan clusters above the claimed 1.3×$10^{-12}$ erg/$cm^{2}$ s completeness limit would show the catalog is not the unbiased census it claims to be.","supporting_citations":[{"cited_title":"& Rodighiero , G","cited_arxiv_id":null,"evidence_quote":"Provides the alternative EBL model that produces stronger spectral hardening for high-redshift sources and the z≈1 optical-depth benchmark."},{"cited_title":"M., & Vovk , I","cited_arxiv_id":null,"evidence_quote":"Establishes the earlier all-sky Fermi-LAT search method and source list that this work extends with sixteen-year exposure."},{"cited_title":"S., Romani, R","cited_arxiv_id":null,"evidence_quote":"Provides the spectroscopic redshift limits used to assess the disputed redshifts of the high-z VHE sources."}],"review_version":1}