{"id":"d4f4fdd3-8c68-4ce3-95cb-d9f1286edbe7","arxiv_id":"2505.09181","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"The authors present 143 giant radio galaxies larger than 3 Mpc, including 69 new discoveries and six above 5 Mpc, and show they are statistically indistinguishable from smaller giants except for hints of smaller bending angles.","lead":"This paper assembles the largest catalog of giant radio galaxies larger than 3 megaparsecs: 143 objects, 69 of them newly identified. It tests whether these extreme sources differ from smaller giants in power, shape, cluster environment, and bending angle. A generalist reader might care because these jets are among the largest structures produced by supermassive black holes and constrain how such jets survive for tens of millions of years.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Photometric/estimated host redshifts and alternative host identifications control sample membership near the 3 Mpc threshold; without spectroscopic confirmation the 'indistinguishable from smaller GRGs' claim is not yet supported.","rationale":"The reader's weakest assumption identifies exactly the same load-bearing point, and I agree with it. I considered the lack of a formal selection function as an alternative concern: the sample is a literature compilation plus new visual discoveries, so population-level fractions and trends are subject to unknown completeness. That is a real limitation, but the paper's most prominent claim is the 'indistinguishable' comparison, and that claim fails specifically if the >3 Mpc threshold membership is unstable. The redshift issue is more load-bearing because it is upstream: it controls which 143 sources enter the sample and, for the six record sources, whether the extreme tail exists at all. The paper is transparent about the problem, even providing lower limits and alternative hosts, but transparency does not remove the dependence of the statistical claims on unmeasured quantities. The proposed spectroscopic test is expensive but decisive; a photo-z propagation test is a cheaper interim check. If the test shows the sample is stable, the conditional verdict can later be upgraded; if not, the statistical conclusions should be downgraded to upper/lower-limit statements. The verdict remains CONDITIONAL because the catalog itself has value as a candidate list, while the population-level claims need the additional support described above.","tokens_in":28806,"tokens_out":5426,"duration_ms":54728,"concrete_test":"Obtain optical spectra for all hosts with redshift type 'p' or 'e' in Table 1 (or, at minimum, the ~18 sources with 3.0 < LLS < 3.5 Mpc plus the six sources with LLS > 5 Mpc), and recompute the LLS with the measured redshifts, including the alternative host identifications already noted in the footnotes. Then re-derive the median redshift, median log P_145, and quasar fraction of the surviving >3 Mpc sample. If more than ~10 per cent of the current 143 sources drop below 3 Mpc, or if the median values shift outside the range spanned by the two comparison samples, the 'indistinguishable' conclusion does not survive. A cheaper preliminary check: propagate the full photo-z PDFs rather than the adopted averages and count how many sources cross 3 Mpc at the 1-sigma lower bound.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that GRGs >3 Mpc are indistinguishable from smaller GRGs in redshift and luminosity depends on the reliability of the 67 per cent of redshifts that are photometric or author estimates. The 3 Mpc threshold is not physical, so modest redshift errors move sources across it. The paper's own footnotes give alternative hosts with very different LLS: J0101+5052 falls from 3.13 to 1.32 Mpc, J0843+0208 from 4.83 to 2.7 Mpc, and J1558-2138 from 3.70 to 0.87 Mpc. All six >5 Mpc sources are quoted as lower limits 'assuming the lowest reasonable host redshift', so the extreme tail of the size distribution is not measured, only bounded. The P-D comparison is similarly unquantified: the authors state that total flux densities can be uncertain by >50 per cent and refrain from quoting errors. If a moderate fraction of the photometric redshifts is overestimated, the sample is contaminated with sub-3-Mpc sources and the comparison against 1-3 Mpc samples is biased toward the null. This is not a claim that the authors are wrong; it is a claim that the central statistical statement is not yet supported at the precision asserted, because sample membership and the compared quantities both depend on unquantified redshifts and fluxes.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper compiles a sample of 143 giant radio galaxies with projected largest linear sizes above 3 Mpc, of which 69 are claimed as new discoveries from LOFAR, ASKAP, MeerKAT, and other surveys. The authors revise hosts and redshifts from the literature, add photometric and spectroscopic redshifts, and measure angular sizes, 145-MHz radio powers, arm-length ratios, bending angles, and cluster associations. Their headline results are that GRGs larger than 3 Mpc are statistically indistinguishable from smaller GRGs in redshift, radio luminosity, and quasar fraction; that they are almost exclusively FR II; that about a quarter show remnant-like diffuse lobes; that at least about 16 per cent are in clusters; and that the bending angle may decrease with size. The paper's main deliverable is the curated catalog in Table 1.","tokens_in":29004,"tokens_out":6927,"duration_ms":64704,"significance":"If the sample and measurements are reliable, this is the largest published census of the extreme tail of the GRG size distribution and a useful reference for jet-environment models; the six sources above 5 Mpc, including Porphyrion, are individually important. The authors are transparent about their methods and flag many host ambiguities in Table 1 footnotes, and they provide the full sample table with provenance codes, which is a practical asset. However, the headline null result is currently fragile because it rests on unquantified photometric redshifts and hand-measured fluxes, and because several record-size sources are lower limits. The paper does not ship machine-checked code or a formal statistical derivation, so its value is empirical; the catalog will likely be used widely once the robustness issues are addressed.","major_comments":[{"comment":"The central claim that GRGs larger than 3 Mpc are indistinguishable from smaller GRGs in redshift and luminosity depends on sample membership, which in turn depends on the 67 per cent of redshifts that are photometric or estimated. The paper itself documents alternative hosts that change the LLS dramatically: J0101+5052 from 3.13 to 1.32 Mpc, J0843+0208 from 4.83 to 2.7 Mpc, and J1558−2138 from 3.70 to 0.87 Mpc (Table 1 footnote a and Section 3.1). Since the 3 Mpc threshold is explicitly not physical, modest redshift or host errors move sources across it and bias the comparison against 1–3 Mpc samples. Please repeat the redshift and luminosity comparisons after (i) excluding all sources with a plausible alternative host and (ii) conservatively assigning the lowest available host redshift, and report how many sources remain above 3 Mpc in each case.","section":"Section 2 and Table 1 (footnotes a–h)"},{"comment":"Radio powers are integrated by hand-drawn Aladin regions, and the text states that total flux densities may be uncertain by more than 50 per cent and that the authors refrain from quoting quantitative error values. The statement that the P–D distribution of the >3 Mpc sample is indistinguishable from smaller GRGs is therefore not quantitatively supported: the comparison has no error bars on either axis, and the spectral index is fixed to alpha = −0.8 for all sources when converting to 145 MHz. Please add at least a systematic flux uncertainty estimate, for example by comparing independent survey measurements for a subset, propagate it to log P, and re-run the distributional comparisons with Monte Carlo realizations of the fluxes and redshifts.","section":"Section 3.1 (P–D diagram)"},{"comment":"Most of the six objects presented as GRGs larger than 5 Mpc have LLS values quoted as lower limits in the text and in the Fig. 4 caption, with the caption specifying that the limits arise from assuming the lowest reasonable host redshift. The abstract's statement that the sample includes GRGs 'clearly exceeding 5 Mpc and reaching up to 6.6 Mpc' therefore overstates what is measured: the extreme tail of the size distribution is bounded, not measured. Please rephrase these statements and, in the statistical analyses, treat the >5 Mpc LLS values as censored data rather than as point measurements.","section":"Section 3.2 and Abstract"},{"comment":"The cluster-association fraction contains an arithmetic inconsistency. The text reads: 'There are 25 of our 143 GRGs with z>=0.9, of which three lie at Galactic latitude |b|<=20 degrees, which leaves us with 124-22 = 102 of the 143 GRGs'. Subtracting 25 from 143 gives 118, and subtracting the three low-latitude sources gives 115; the expression 124−22=102 is not explained. Since the conclusion that at least 16 per cent of GRGs are in clusters depends on the denominator, please correct the counting and recompute the fraction and its uncertainty.","section":"Section 3.4"}],"minor_comments":[{"comment":"The provenance code description says 'MA = MeerKAT MALS DR2 ... at 1.27 MHz'; this should read 1.27 GHz, as the MALS survey is an L-band survey.","section":"Table 1 caption"},{"comment":"J0740−6647 is listed among the nine sources that deviate from FR II, but the text says its morphology 'formally still conforms to an FR II type on both sides'; please clarify the criterion for inclusion in that list.","section":"Section 3.3"},{"comment":"The environment-flag discussion says one GRG host is listed as 'bi' but later says 'three further GRGs are even listed as bi'; please align the text with the flags in Table 1, which appear to include more than one 'bi' entry.","section":"Section 3.4"},{"comment":"The abstract's 'tentative evidence that the bending angle decreases with size' is appropriately cautious, but the text's first comparison (median 3.0 degrees versus 1.4 degrees for LLS below and above 4 Mpc) is followed by equal-size splits with p = 0.186 and p = 0.88; please present the primary statistical test consistently in the abstract and conclusions.","section":"Section 3.5"}],"recommendation":"major_revision","confidential_remarks":"The paper is a useful curated catalog, but the headline 'indistinguishable' claim is not yet supported at the precision asserted because the underlying redshifts, fluxes, and sample membership are unquantified. The issues are fixable with robustness checks, so I recommend major revision rather than rejection. I would also ask the editor to ensure that the '69 newly found' count is clearly delimited against overlapping recent catalogs such as Mostert et al. (2024) and Koribalski (2025), since the provenance code alone is not a formal novelty disclosure."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This is a catalog paper, and judged as a catalog it is genuinely useful. The new sample of 143 GRGs larger than 3 Mpc, with 69 new discoveries and six above 5 Mpc, is exactly what the field needs for studying the extreme end of the size distribution. The authors also do real service by revising host identifications for several previously published sources, and they are admirably transparent about cases where an alternative host would change the size dramatically (J0101+5052, J1558-2138, J0843+0208). The morphological statistics—mostly FR II, a quarter with remnant-like diffuse lobes, only 59% with hotspots—are new for this population and look robust to the redshift issues.\n\nThe soft spots are real but not fatal. Roughly two-thirds of the redshifts are photometric or author estimates, and the 3 Mpc threshold is not physical, so modest redshift errors move sources in and out of the sample. The paper's own footnotes show that plausible alternative hosts would drop J0101+5052 from 3.13 to 1.32 Mpc and J1558-2138 from 3.70 to 0.87 Mpc. That directly weakens the central claim that this population is indistinguishable from smaller GRGs in redshift and luminosity: if a chunk of the sample is actually sub-3-Mpc sources, the comparison is biased toward the null. The flux integration is hand-drawn in Aladin with no quoted errors, and the authors say uncertainties can exceed 50%, so the P-D outliers should be treated as tentative. I also missed a machine-readable table in the arXiv version; for a catalog paper that is close to a requirement.\n\nThe comparison samples from Andernach et al. and Simonte et al. are partly built by the same group, but that is not a real flaw here—the samples are what they are, and the authors are open about the heterogeneity. The paper does not oversell its conclusions; the abstract and discussion carefully hedge the bending-angle and cluster-fraction trends, and the six >5 Mpc sizes are explicitly lower limits. On balance, the central statistical statement is plausible but not yet pinned down at the precision asserted. That is a reason to demand a data release and error estimates, not a reason to reject the catalog.\n\nI would send this to a serious referee. The sample is important, the work is careful, and the weaknesses are addressable with a revised manuscript plus supplementary tables. Reading-group value is moderate: good for a discussion of how selection effects and redshift uncertainties shape extreme-object samples.","headline":"A useful and honest catalog of the most extreme giant radio galaxies; the sample itself is the contribution, while the statistical null results rest on photometric redshifts and hand-drawn fluxes that need to be published with error bars before being taken as firm.","tokens_in":29624,"tokens_out":1356,"would_cite":true,"duration_ms":15045,"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":"The paper builds a sample of 143 giant radio galaxies larger than 3 Mpc and claims they are statistically indistinguishable from smaller giant radio galaxies in redshift, radio power, and quasar fraction, while being almost exclusively FR…","keywords":["giant radio galaxies","radio jets","FR II morphology","photometric redshifts","radio power-size relation","bending angles","galaxy clusters","LOFAR surveys"],"falsifier":"Obtain spectroscopic redshifts for a random subset of about 30 of the roughly 95 hosts that currently have only photometric or estimated redshifts, then re-derive the projected linear sizes. If the new redshifts are systematically lower and push a significant fraction of the 143 sources below 3 Mpc, the claim of indistinguishability and the six >5 Mpc record sizes would collapse; confirmation would solidify the sample. A quicker check is to test the flagged alternative hosts, for example J0101+5052, whose alternative host at $z=0.225$ would reduce its size to 1.32 Mpc.","tokens_in":28518,"feed_emoji":"📡","tokens_out":5526,"duration_ms":51415,"temperature":0.7,"pith_summary":"The paper assembles, from published catalogues and its own visual search of modern radio surveys, a sample of 143 giant radio galaxies whose projected linear size exceeds 3 Mpc, 69 of which are newly identified. It asks whether these extreme sources differ from the much larger population of giant radio galaxies above 0.7 Mpc. The answer, on the paper's evidence, is mostly no: the >3 Mpc sources match smaller giants in median redshift, radio luminosity, quasar fraction, and cluster association, and are almost entirely FR II-type radio galaxies. The sample includes six sources larger than 5 Mpc, the largest reaching roughly 6.6 Mpc, whose straightness and low bending angles raise puzzles for jet-formation models. The paper also reports tentative evidence that bending angle decreases with size and that cluster-associated giants have larger bending angles.","feed_headline":"Giant radio galaxies over 3 Mpc look just like smaller ones","feed_subtitle":"A 143-source census, 69 newly found, shows the biggest radio galaxies are not special in power, redshift, or quasar fraction.","key_machinery":"The central object is the curated sample itself: 143 sources with measured largest angular size, host redshift (33 per cent spectroscopic, 67 per cent photometric or estimated), projected linear size, radio power at 145 MHz extrapolated with spectral index $-0.8$, arm-length ratio (brighter-to-fainter lobe), bending angle, and a cluster-environment flag. The argument is carried by comparisons of this sample's distributions with those of smaller giants (1--3 Mpc) in redshift, radio power, quasar fraction, and cluster association, plus internal trends of bending angle and lobe asymmetry. The power--size diagram (radio power at 145 MHz versus linear size) is used to identify outliers that challenge evolutionary models of radio galaxies.","core_discovery":"The paper claims that the extreme tail of the giant radio galaxy population is continuous with the rest: giants larger than 3 Mpc are drawn from the same parent population as giants between 0.7 and 3 Mpc, with statistically indistinguishable median redshift, radio luminosity, quasar fraction, and cluster association fraction, while being near-universally of FR II morphology (at most one clear FR I) and often showing diffuse, remnant-like lobes. Six sources exceed 5 Mpc in projected size; for most of these the quoted size is a lower limit set by the lowest plausible host redshift. The authors interpret the lack of distinguishing features as evidence that extreme size does not require a special environment or jet mechanism, but rather that the largest sources are simply the long-lived, straight, and rare tail of the giant radio galaxy distribution.","pith_inferences":["The statistical indistinguishability claim rests heavily on photometric redshifts for two-thirds of the sample; if those redshifts carry systematic biases, the >3 Mpc selection could be contaminated by smaller sources and the 'no difference' result could be an artefact. A spectroscopic follow-up of a few dozen hosts would settle this.","The record sizes above 5 Mpc depend on single host identifications; as the authors themselves note for J0101+5052 and J1558−2138, an alternative host choice can shrink the linear size by factors of two or more, so the true maximum size of radio galaxies remains uncertain.","The steep drop-off in counts beyond 3 Mpc, if confirmed with better redshifts, could be combined with models of jet power and source age to constrain the maximum jet lifetime and the magnetic-field seeding of cosmic voids.","The outliers in the power--size diagram—both the four overluminous and the two underluminous sources—are natural laboratories for testing whether standard radio-galaxy evolution models need additional ingredients such as re-acceleration or intermittent jet activity."],"forward_implications":["If the sample is representative, the number of giants larger than 3 Mpc drops steeply with size (cumulative slope of $-6$ or steeper), making these sources rare probes of jet longevity and intergalactic medium properties.","The near-total absence of FR I morphology above 3 Mpc implies that the FR I/II division persists even at extreme physical scales.","The cluster association fraction of at least 16 per cent, including several brightest cluster galaxies, argues that underdense environments are not required to build Mpc-scale radio sources; high jet power may suffice.","The six straight sources larger than 5 Mpc, with bending angles at most $3.8\\degree$, constrain jet stability and imply host-galaxy peculiar velocities below roughly $10^2$ km/s.","The tentative decrease of bending angle with linear size, if real, suggests that longer jets are straighter, possibly because they grow into lower-density media."],"supporting_citations":[{"why":"Defines the giant radio galaxy class with the 0.7 Mpc size threshold that the paper builds on.","marker":"Willis et al. 1974"},{"why":"Supplies a large published compilation of giants that feeds the >3 Mpc sample.","marker":"Kuźmicz et al. 2018"},{"why":"Provides the 180-source 1--3 Mpc comparison sample used for median redshift, power, and quasar fraction baselines.","marker":"Andernach et al. 2021"},{"why":"Gives the LoTSS deep-field sample of 128 giants in the 1--3 Mpc range, a second comparison set, and a source of photometric redshifts.","marker":"Simonte et al. 2024"},{"why":"Supplies a large LoTSS-discovered giant sample, the Pareto tail index for comparison, and several specific sources including Alcyoneus.","marker":"Oei et al. 2023a"},{"why":"The LoTSS DR2 paper from which three of the six >5 Mpc sources and many host identifications are taken.","marker":"Hardcastle et al. 2023"},{"why":"Provides the spectroscopic redshift and physical analysis of J1529+6015 (Porphyrion), the second-largest source in the sample.","marker":"Oei et al. 2024"},{"why":"The LoTSS DR2 survey images used for discovery, size measurement, and flux integration for most of the northern sources.","marker":"Shimwell et al. 2022"},{"why":"The RACS survey images used to find and measure several of the new giants and to check host identifications.","marker":"McConnell et al. 2020"},{"why":"A photometric redshift catalogue that supplies redshifts for many of the hosts lacking spectroscopy.","marker":"Beck et al. 2021"}],"fun_headline_variants":["No special recipe for radio galaxies over 3 Mpc","Biggest radio galaxies fit the same mold as smaller ones","Extreme-size radio galaxies: same population, longer reach","143 giants, 69 new: size alone doesn't set them apart"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The sample's sizes and statistics rest on photometric or estimated redshifts for two-thirds of the hosts, and on the authors' choice of host galaxy for several of the record-size sources; if a substantial fraction of these are wrong, the >3 Mpc selection and the size-dependent trends would change.","fun_headline_variants_meta":{"raw":{"variants":["No special recipe for radio galaxies over 3 Mpc","Biggest radio galaxies fit the same mold as smaller ones","Extreme-size radio galaxies: same population, longer reach","143 giants, 69 new: size alone doesn't set them apart"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000675,"raw_usage":{"total_tokens":3145,"prompt_tokens":1095,"completion_tokens":2050,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":711,"completion_tokens_details":{"reasoning_tokens":1980}},"tokens_in":711,"tokens_out":2050,"duration_ms":16214,"temperature":1.0,"reasoning_tokens":1980,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T21:36:50.931784+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Obtain spectroscopic redshifts for a random subset of about 30 of the roughly 95 hosts that currently have only photometric or estimated redshifts, then re-derive the projected linear sizes. If the new redshifts are systematically lower and push a significant fraction of the 143 sources below 3 Mpc, the claim of indistinguishability and the six >5 Mpc record sizes would collapse; confirmation would solidify the sample. A quicker check is to test the flagged alternative hosts, for example J0101+5052, whose alternative host at $z=0.225$ would reduce its size to 1.32 Mpc.","supporting_citations":[{"cited_title":"J., Horton , M","cited_arxiv_id":null,"evidence_quote":"The LoTSS DR2 paper from which three of the six >5 Mpc sources and many host identifications are taken."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the spectroscopic redshift and physical analysis of J1529+6015 (Porphyrion), the second-largest source in the sample."}],"review_version":1}