{"id":"f324630f-7d90-425d-9773-4113f8006448","arxiv_id":"1908.05137","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"IceCube's stacking analysis for 189 ULIRGs would need roughly 10 to 100 neutrinos, depending on spectral index, to reach its sensitivity.","lead":"This IceCube conference paper selects 189 nearby Ultra-Luminous Infrared Galaxies and computes the sensitivity of an 8-year stacked neutrino search for them. It asks whether these dusty, gamma-ray-dark galaxies could be the hidden sources of the cosmic neutrinos IceCube sees.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Standard-candle weighting, not catalog bias, is the load-bearing assumption behind the 10-100 neutrino sensitivity claim.","rationale":"The reader identified the standard-candle assumption as the weakest point, and I agree: the weighting Wk = dk^{-2} is the necessary link between the simulated signal and the physical source population. The paper is transparent about this assumption, and it labels the results as preliminary sensitivities, so the argument is internally consistent. However, the load-bearing nature of the assumption is real and testable: a reweighting by L_IR or by AGN fraction could shift the quoted 10-100 neutrino range. The reader's conditional verdict already asks for this assumption to be tested or relaxed, so no change to the verdict is needed. I did not identify a more serious flaw: the sensitivity calculation follows the standard IceCube unbinned stacking formalism, the energy range is convolved with the effective area, and the pseudo-experiment procedure is appropriate. The catalog-bias issue noted by the reader is a secondary concern about the interpretation of future upper limits, not about the sensitivity curves themselves. Thus my read confirms the conditional verdict.","tokens_in":6147,"tokens_out":6180,"duration_ms":65260,"concrete_test":"Run a mismatched-model pseudo-experiment: inject signal events using weights Wk = L_IR,k dk^{-2} (or a two-component model where, e.g., the 20 AGN-dominated sources produce 90% of the flux), but keep the likelihood signal PDF fixed at the standard-candle weights Wk = dk^{-2}. Recompute the 90% sensitivity and 5-sigma discovery potential for gamma in [1.5, 3.5]. If the required total number of detected neutrinos n_nu moves outside the quoted 10-100 range for any spectral index, the standard-candle assumption is load-bearing; if the shift is small, the quoted sensitivity is robust to this misspecification.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The headline sensitivity (10-100 neutrinos, abstract, Fig. 4, and conclusions) is obtained from the stacked signal PDF in Eq. 3.1, where each ULIRG contributes with weight Wk = dk^{-2} under the explicit assumption that all 189 sources are neutrino standard candles (Sec. 3.1). This weighting is the physical map from the pseudo-experiment signal injection to the real ULIRG population. If the true neutrino luminosity instead scales with L_IR, which spans 10^12 to 10^13 L_sun in Fig. 3, or is concentrated in a minority of AGN-hosting systems, then the true spatial signal distribution differs from the assumed S(x_i, gamma). The quoted sensitivity is then a sensitivity to the standard-candle model, not to ULIRGs as a class, and future upper limits on the cumulative neutrino flux of the 189 ULIRGs would be difficult to interpret. The paper's claim that the sample is unbiased by construction does not mitigate this: unbiasedness of source selection concerns which sources are included, not how their individual neutrino fluxes are weighted. The assumption is stated transparently, so the paper is not misleading, but the central claim inherits the assumption. The right-panel conversion from flux normalization to n_nu absorbs the same weighting, so the 10-100 neutrino range is not robust to alternative, physically motivated weighting schemes.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper motivates Ultra-Luminous Infrared Galaxies (ULIRGs) as candidate gamma-ray obscured sources of the IceCube diffuse neutrino flux. It constructs a sample of 189 ULIRGs from three IRAS-based catalogs, characterizes their redshift and infrared-luminosity distributions, and presents preliminary sensitivities for an 8-year IceCube time-integrated stacking analysis. Using an unbinned likelihood with a stacked signal PDF weighted by W_k = d_k^{-2} under an explicit neutrino standard-candle assumption, the authors report that the analysis is sensitive to roughly 10 to 100 neutrinos in the 10 GeV to 70 PeV range, depending on the assumed common spectral index. The paper argues that a non-detection would produce the first limits on the cumulative high-energy neutrino flux from ULIRGs and would constrain hadronic pp-interaction models in these dust-rich environments.","tokens_in":6324,"tokens_out":5438,"duration_ms":53243,"significance":"This is a useful first step: the paper presents the first IceCube source selection and sensitivity study for ULIRGs as a source class, and the methodological components—likelihood, test statistic, pseudo-experiments, signal injection, and detector-response convolution—are described clearly enough to be reproduced. The standard-candle weighting assumption and the common-spectral-index assumption are stated transparently, which is a strength. If the sensitivity estimates are confirmed by the subsequent analysis, a non-detection would yield meaningful constraints on pp-dominated hadronic emission in ULIRGs and would test the gamma-ray-obscured source scenario for the diffuse neutrino flux.","major_comments":[{"comment":"The headline sensitivity of 10 to 100 neutrinos is obtained with weights W_k = d_k^{-2}, i.e., under the assumption that all 189 ULIRGs are neutrino standard candles. This weighting directly determines the stacked signal PDF in Eq. (3.1) and the conversion between flux normalization and n_nu in the right panel of Fig. 4. If the true neutrino luminosity scales with L_IR or is dominated by a minority of AGN-hosting systems, the quoted sensitivity describes a different source population and the 10-100 neutrino numbers are not robust. I recommend adding a robustness check with an alternative physically motivated weighting (for example W_k proportional to L_{IR,k} d_k^{-2}) for at least one spectral index, or explicitly stating in the abstract and conclusions that the sensitivity is conditional on the standard-candle model.","section":"Section 3.1, Eq. (3.1) and Fig. 4"},{"comment":"The claim that the selected sample is 'by construction unbiased' is internally inconsistent with the description of the FSC catalog, whose ULIRGs require an available optical counterpart for redshift measurement. This introduces a selection effect that could bias the sample toward optically brighter or nearer sources. Because Section 4 repeats the unbiasedness claim as a strength of the sample, the authors should either justify it quantitatively with a comparison to the full IRAS flux-limited parent sample or soften the wording to 'flux-limited without full-sky completeness.'","section":"Section 2"}],"minor_comments":[{"comment":"The sensitivity definition, 'the amount of injected signal required such that in 90% of the pseudo-experiments one obtains a p-value p <= 0.5,' is nonstandard; please clarify how it relates to the commonly used median 90% confidence-level sensitivity, since a reader may otherwise misinterpret the quoted 10-100 neutrino range as a standard 90% upper limit.","section":"Section 3.2"},{"comment":"The text 'the stacking is performed in the form of a weighed sum' contains a typo; it should read 'weighted sum.'","section":"Section 3.1"},{"comment":"The right-panel label 'Total number of ULIRG neutrinos' should specify that this is the number of neutrino events expected in the detector after convolution with the effective area, not the number emitted by the sources.","section":"Section 3.2 and Fig. 4"},{"comment":"Several reference entries contain LaTeX spacing artifacts, such as 'F ermi-LA TCollaboration' and 'F ermi-LA T'; these should be corrected to 'Fermi-LAT Collaboration' and 'Fermi-LAT.'","section":"References"},{"comment":"The introduction mentions both p-gamma and pp channels, while the conclusions refer only to the pp-interaction channel; please specify that the pp channel is the dominant production mechanism in the dust-rich star-forming regions considered here, or otherwise keep the channel discussion consistent.","section":"Conclusions"}],"recommendation":"major_revision","confidential_remarks":"This is a conference proceedings paper with preliminary results. The main technical concern is the standard-candle weighting assumption, which is stated transparently and is therefore not misleading, but it should be stress-tested or explicitly qualified given that the abstract and conclusions quote the 10-100 neutrino range without this caveat. The 'unbiased' claim in Section 2 is a factual overstatement that should be corrected. With those revisions, the paper would be appropriate for the proceedings."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The short version: this is a clean, honest sensitivity study for a genuinely new target class. It does not claim a detection; it builds a 189-source ULIRG catalog and shows that an 8-year IceCube stacking analysis would be sensitive to 10-100 neutrinos for spectral indices between 1.5 and 3.5. The new thing is the catalog and the sensitivity curves; the likelihood stacking itself is Braun et al. 2008. That is fine--novelty in application is enough.\n\nWhat it does well: the source selection is transparent, from three IRAS catalogs with duplicate matching, coordinates and redshifts from NED, and the ULIRG properties are plotted. The sensitivity calculation is specified: unbinned likelihood, signal PDF with detector weights and d^-2 distance weights, pseudo-experiments, definitions of sensitivity and discovery potential, and an equation for n_nu. The assumptions are stated rather than hidden. The paper would be reproducible from the text.\n\nWhere I have reservations:\n\nFirst, and this is the load-bearing one, the sensitivity to 10-100 neutrinos is computed under the assumption that all 189 ULIRGs are neutrino standard candles. The weighting W = d^-2 is the physical map from the injected signal to the real population. If neutrino luminosity instead tracks L_IR, or is dominated by a handful of AGN-hosting systems, the signal PDF is wrong and the quoted sensitivity is sensitivity to the model, not to ULIRGs as a class. The assumption is explicitly stated in Sec. 3.1, so the paper is not misleading, but the headline number inherits it. The authors should at least run a luminosity-weighted version or show the effect on the curves; for a first feasibility study I would accept the standard-candle scenario as a default, but not as the only one.\n\nSecond, 'unbiased by construction' is overstated. The catalogs are flux-limited, and the FSC sample requires optical counterparts for redshifts. The selection may be fine for a first search, but calling it unbiased without caveat goes beyond what the text shows.\n\nThird, the sensitivity curves carry no systematic uncertainties--no IceCube effective-area systematics, no atmospheric-background modeling uncertainty. For a conference proceeding that is acceptable, but any full paper should include error bands.\n\nThe circularity burden is low: no data are fitted, the source list is independent of IceCube, and self-citations to IceCube detector papers are appropriate background. The motivation via gamma-ray opacity is physically sensible and the paper does not oversell the interpretation.\n\nWho should read this: anyone preparing a stacking search for steady extragalactic source classes, and any theorist who models ULIRG neutrino emission. It deserves conference presentation and, if expanded into a full paper, a serious referee. My advice: engage with it, but treat the 10-100 neutrino number as conditional on the standard-candle assumption, not as a robust statement about ULIRGs.","headline":"A clean sensitivity study for a new target class, but the 10-100 neutrino headline is conditional on treating 189 ULIRGs as standard candles.","tokens_in":6906,"tokens_out":3619,"would_cite":true,"duration_ms":38757,"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 proposes that ULIRGs, the infrared-brightest galaxies, are prime candidates for the gamma-ray-hidden sources of IceCube's high-energy neutrinos.","keywords":["ultra-luminous infrared galaxies","high-energy neutrinos","IceCube","stacking analysis","gamma-ray obscured sources","hadronic acceleration","pp interactions","diffuse neutrino flux"],"falsifier":"The decisive check is the stacking search itself: if 8 years of IceCube data on the 189 selected ULIRGs yield a best-fit signal consistent with zero and a 90% upper limit below the sensitivity shown for gamma = 2.0, then the ULIRG population cannot be hiding a major share of the diffuse neutrino flux.","tokens_in":5914,"feed_emoji":"🔭","tokens_out":6705,"duration_ms":64601,"temperature":0.7,"pith_summary":"The paper sets out to establish that ultra-luminous infrared galaxies (ULIRGs) are prime candidate sources of the high-energy neutrinos detected by the IceCube observatory, and that they can account for the apparent mismatch between the neutrino flux and the diffuse gamma-ray flux. The argument is that these galaxies contain both the hadronic accelerators (starburst regions and AGN) and the dust needed to absorb the gamma rays that would otherwise accompany neutrino production. To make the case testable, the paper assembles an unbiased sample of 189 nearby ULIRGs and shows that a stacking analysis of eight years of IceCube data would be sensitive to a total of 10 to 100 neutrinos in the energy range from about 10 GeV to 70 PeV, depending on the spectral index. If real, the result would either identify ULIRGs as a new class of neutrino sources or set the first upper limits on their neutrino flux.","feed_headline":"189 dusty galaxies could solve the missing-gamma-ray puzzle","feed_subtitle":"Stacking eight years of IceCube data can reveal whether ULIRGs are gamma-ray-hidden neutrino sources.","key_machinery":"The central mechanism is a time-integrated unbinned maximum-likelihood stacking analysis. Each of the 189 ULIRGs contributes a signal probability density function with a spatial and an energy component; the per-source PDFs are combined in a weighted sum, with detector weights R_k and theoretical weights W_k = $d_k^{{-2}}$ assigned from comoving distance under the assumption that ULIRGs are neutrino standard candles. The signal energy PDF depends on a common spectral index gamma, and the sensitivity and discovery potential are computed by injecting power-law pseudo-signals and scrambling data in right ascension to build the background test-statistic distribution.","core_discovery":"The central claim is that ULIRGs are ideal gamma-ray obscured neutrino source candidates: they combine hadronic acceleration sites (starbursts and AGN) with abundant dust that both provides target material for pp neutrino production and absorbs the accompanying gamma rays. If this is right, the apparent tension between the diffuse neutrino flux measured by IceCube and the diffuse gamma-ray flux measured by Fermi-LAT disappears, because the gamma rays are hidden at the source. The paper demonstrates the first practical test: a stacking analysis of 189 unbiased, nearby ULIRGs with 8 years of IceCube data, whose sensitivity lies between 10 and 100 neutrinos in the energy range 10 GeV to 70 PeV depending on the spectral index.","pith_inferences":["If neutrino luminosity actually scales with infrared luminosity instead of being identical across sources, a stacking analysis weighted by L_IR could be more sensitive than the equal-luminosity weighting used here; the paper's own luminosity distribution provides the input for such a test.","A detection of neutrinos from these z < 0.35 galaxies would make ULIRGs a concrete contributor to the diffuse IceCube flux, and the same dust that absorbs gamma rays would be expected to reprocess the energy into a lower-energy electromagnetic cascade that future gamma-ray instruments could search for independently.","If ULIRGs show no signal, the constraints would generalize to the broader class of infrared-luminous dusty galaxies, making them a natural next frontier for hidden-source searches."],"forward_implications":["A stacking analysis of 8 years of IceCube data can detect a cumulative neutrino signal from the 189 ULIRGs even if each individual galaxy is too faint to be seen on its own.","A detection would identify ULIRGs as a new class of high-energy neutrino sources and would support the gamma-ray-obscured resolution of the IceCube and Fermi-LAT tension.","A null result would set the first upper limits on the cumulative neutrino flux from ULIRGs, constraining hadronic acceleration and pp-channel production in these systems.","The required number of neutrinos ranges from roughly 10 for hard spectra to around 100 for soft spectra, meaning the reach of the analysis depends strongly on the unknown spectral index gamma.","Because the 189-object sample is unbiased and all-sky, the result applies to the ULIRG population as a whole rather than to a few exceptional sources."],"supporting_citations":[{"why":"Introduces gamma-ray obscured sources as a way to relieve the tension between the diffuse neutrino and gamma-ray fluxes.","marker":"[8]"},{"why":"Shows that the non-blazar gamma-ray flux underestimates the IceCube neutrino flux, establishing the tension that motivates hidden sources.","marker":"[7]"},{"why":"Reports the first observation of a diffuse high-energy cosmic neutrino flux with IceCube.","marker":"[1]"},{"why":"Provides the review of ULIRGs, including their infrared luminosities, starburst rates, and dusty environments.","marker":"[9]"},{"why":"Supplies evidence for AGN in ULIRGs and the third catalog of ULIRGs from PSCz and Spitzer observations.","marker":"[12]"},{"why":"The IRAS Revised Bright Galaxy Sample, the first catalog used for the ULIRG selection and the preferred source of infrared luminosity values.","marker":"[15]"},{"why":"The IRAS ULIRG survey from the Faint Source Catalog, the second catalog contributing 118 sources to the selection.","marker":"[16]"},{"why":"Establishes the unbinned maximum-likelihood stacking formalism that the analysis uses to combine the 189 ULIRG signal PDFs.","marker":"[19]"},{"why":"Provides the Planck 2015 cosmology used to convert redshifts into the comoving distances that set the d^{-2} weights.","marker":"[20]"}],"fun_headline_variants":["Dusty ULIRGs could hide gamma rays and shine in neutrinos","189 dusty galaxies might be hidden neutrino sources","ULIRGs: gamma-ray-hidden sites for IceCube neutrinos","Could ULIRGs crack the missing gamma-ray puzzle?","Stacking IceCube data to test dusty neutrino factories"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"All 189 ULIRGs are treated as identical neutrino emitters, so their contributions are weighted only by distance ($d^{{-2}}$); if neutrino output actually varies from galaxy to galaxy, the computed sensitivities will not describe the real population.","fun_headline_variants_meta":{"raw":{"variants":["Dusty ULIRGs could hide gamma rays and shine in neutrinos","189 dusty galaxies might be hidden neutrino sources","ULIRGs: gamma-ray-hidden sites for IceCube neutrinos","Could ULIRGs crack the missing gamma-ray puzzle?","Stacking IceCube data to test dusty neutrino factories"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000677,"raw_usage":{"total_tokens":3041,"prompt_tokens":867,"completion_tokens":2174,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":483,"completion_tokens_details":{"reasoning_tokens":2091}},"tokens_in":483,"tokens_out":2174,"duration_ms":14054,"temperature":1.0,"reasoning_tokens":2091,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:21:52.881152+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"The decisive check is the stacking search itself: if 8 years of IceCube data on the 189 selected ULIRGs yield a best-fit signal consistent with zero and a 90% upper limit below the sensitivity shown for gamma = 2.0, then the ULIRG population cannot be hiding a major share of the diffuse neutrino flux.","supporting_citations":[{"cited_title":"Bechtol et al., The Astrophysical Journal 836 (2017) 47","cited_arxiv_id":null,"evidence_quote":"Shows that the non-blazar gamma-ray flux underestimates the IceCube neutrino flux, establishing the tension that motivates hidden sources."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the review of ULIRGs, including their infrared luminosities, starburst rates, and dusty environments."},{"cited_title":"Nardini et al., MNRAS 405 (2010) 2505–2520","cited_arxiv_id":null,"evidence_quote":"Supplies evidence for AGN in ULIRGs and the third catalog of ULIRGs from PSCz and Spitzer observations."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The IRAS Revised Bright Galaxy Sample, the first catalog used for the ULIRG selection and the preferred source of infrared luminosity values."},{"cited_title":"Kim and D","cited_arxiv_id":null,"evidence_quote":"The IRAS ULIRG survey from the Faint Source Catalog, the second catalog contributing 118 sources to the selection."},{"cited_title":"Braun et al., Astroparticle Physics 29 (2008) 299 – 305","cited_arxiv_id":null,"evidence_quote":"Establishes the unbinned maximum-likelihood stacking formalism that the analysis uses to combine the 189 ULIRG signal PDFs."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the Planck 2015 cosmology used to convert redshifts into the comoving distances that set the d^{-2} weights."}],"review_version":1}