{"id":"1cebf07e-57c4-437f-a642-e7f5d7a21d93","arxiv_id":"2507.06176","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"IceCat-2 reprocesses IceCube alert tracks with a new reconstruction, shrinking 50% localization areas by a factor of about 5 and finding no new catalog coincidences beyond chance.","lead":"IceCube presents a preliminary updated catalog of 365 neutrino alert tracks with roughly five times smaller angular uncertainty, using a new reconstruction method. The improved maps change some source associations, including casting doubt on three tidal disruption events as neutrino counterparts.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The factor-of-5 improvement and the TDE exclusion both depend on unvalidated 50%/90% containment calibration; a simulation-based coverage check on the exact 365-event sample would settle it.","rationale":"The reader's weakest assumption identifies the same load-bearing point: the catalog's improved areas and the TDE exclusion are only as good as the containment calibration. My stress-test adds that the factor-5 improvement is a median ratio of reported areas, not a measured angular-resolution gain; it could reflect a change in how systematics are folded into the contours. Since the catalog is not yet public and the reconstruction papers [9-11] are not reproduced here, the only independent check available to a reader is a simulation-based coverage test. I agree with the CONDITIONAL verdict: the claims are plausible and important but not yet verifiable from this paper. The TDE 'ruling out' would additionally need a fuller multimessenger likelihood treatment including the temporal term, but that concern is secondary to calibration.","tokens_in":10527,"tokens_out":6310,"duration_ms":72044,"concrete_test":"Reconstruct a sample of simulated muon-neutrino events that pass the same alert selection and IceTop veto as the 365 events, using the hybrid algorithm of refs [9-11]. For each event, record whether the true injected direction lies inside the reported 50% and 90% containment regions, binned by energy, declination, and Gold/Bronze channel. If the empirical coverage is consistent with 50%/90% within Poisson errors (e.g., 95% CI) for all bins, the headline improvement is supported; if any bin relevant to the three TDE alerts shows under-coverage (e.g., 90% contours containing only ~70% of true directions), the 'effectively ruling out' claim fails and the catalog needs a caveat.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative claims—median 50%/90% containment areas of 0.4/1.3 square degrees (a factor ~5/4 reduction from IceCat-1) and the statement that TDE positions AT2019dsg, AT2019fdr, AT2019aalc are 'well outside the updated containment regions, effectively ruling out an association'—rest on the assumption that the new hybrid reconstruction's reported containment contours are correctly calibrated for every alert. The paper does not validate this. It refers to refs [9-11] for the algorithm and systematics, but does not show that those calibrations transfer to the 365-event catalog after reprocessing, the IceTop veto cut, and the Gold/Bronze selection. If the 90% contours under-cover for some topologies (e.g., soft-spectrum events with short tracks, or events near the horizon where atmospheric-muon vetoes alter the event sample), a TDE position outside a nominal 90% contour could still be a plausible counterpart, and the claimed 'ruling out' would be unsafe. The 50% containment area alone cannot be checked from the paper because no event-level data or validation plots are given. The text also does not quantify how far outside the contours the TDE positions lie, nor does it combine the spatial exclusion with the original temporal coincidence and a trials factor, so even with calibrated contours the 'ruling out' wording is stronger than the analysis supports.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents IceCat-2, a preliminary update of the IceCube alert-track catalog, containing 365 track-like alerts from May 2011 to January 2025. All events are reprocessed with a newer hybrid reconstruction algorithm, and the paper reports that the median 50% and 90% containment areas are reduced to 0.4 and 1.3 square degrees, respectively, a factor of about 5 (4) relative to IceCat-1. The paper also revisits claimed associations with TXS 0506+056, NGC 7469, and three tidal disruption events (AT2019dsg, AT2019fdr, AT2019aalc), concluding that the TDE positions are 'well outside the updated containment regions, effectively ruling out an association.' Finally, it cross-correlates the 365 alerts with several gamma-ray and X-ray catalogs and finds the number of coincidences consistent with chance expectations based on 1000 right-ascension randomizations.","tokens_in":10754,"tokens_out":4465,"duration_ms":52570,"significance":"If the reported resolution improvement is validated, IceCat-2 would be a valuable public resource for multimessenger follow-up and source-population studies. The paper is commendably transparent about its preliminary status and provides a direct comparison with IceCat-1, including angular separations between old and new reconstructions and the ratio of containment areas. The use of a right-ascension randomization to estimate chance coincidences with external catalogs is a sensible and reproducible null test. However, the two headline claims—the factor-of-5 improvement and the 'ruling out' of TDE associations—depend on the calibration of the 50%/90% containment contours, which is not validated in this manuscript, and the TDE conclusion is drawn from a spatial containment test alone, without the temporal and spectral information that motivated the original associations or a trials-factor treatment. These issues undermine the strength of the conclusions as currently worded, though they are addressable in revision.","major_comments":[{"comment":"The factor-of-5/4 improvement in containment areas and the median 50%/90% areas of 0.4/1.3 square degrees rest on the assumption that the new hybrid reconstruction's containment contours are correctly calibrated for the exact 365-event IceCat-2 sample. The paper refers to refs. [9–11] for the reconstruction and systematics, but it does not show that those calibrations transfer to the catalog after the IceTop veto removes a subset of events, after the Gold/Bronze selection, and after reprocessing of alerts with varying topologies and energies. A simulation-based coverage check on the exact selection—e.g., the fraction of injected signal events falling inside the reported 50%/90% contours as a function of energy, declination, and track topology—would settle whether the reported improvement factors and the TDE exclusion are safe. Without this, the central quantitative claims are unvalidated for the presented catalog.","section":"Sec. 2.1 and Figs. 2–3; Summary"},{"comment":"The statement that the TDE positions are 'well outside the updated containment regions, effectively ruling out an association' is stronger than the analysis supports. The test shown is purely a spatial containment check against the 90% contours; it does not quantify the angular separations in units of the reconstruction uncertainty, does not incorporate the temporal coincidence (~100 days after optical–UV peak) that motivated the original associations in refs. [13–15], and does not account for the trials factor of checking three TDEs that were selected because of prior multimessenger interest. A position outside a nominal 90% spatial contour does not by itself rule out association unless a joint spatial–temporal–spectral likelihood is evaluated with appropriate trial corrections. Please either soften the conclusion to 'not spatially coincident within the 90% containment region' and give the separations, or perform a full multimessenger statistical test and report a trials-corrected p-value.","section":"Sec. 3, TDE paragraph (Fig. 6)"},{"comment":"The expected-coincidence calculation using 1000 right-ascension randomizations is a reasonable null, but the text does not state whether the event-by-event 90% containment contours used for the observed search are also used when scoring the randomized directions, or whether the search is rerun with contours derived from the shuffled directions. If the observed contours are reused, the null distribution is not valid because the contours are tied to the observed best-fit directions and would not correspond to the randomized events. Please clarify the procedure explicitly, including whether the source catalogs are fixed and whether declination is preserved in the randomization.","section":"Sec. 4, Table 1 and accompanying text"}],"minor_comments":[{"comment":"The wording is internally inconsistent: the Abstract says the 50% (90%) angular uncertainty 'is expected to be reduced by a factor of approximately 5(4)', while the Summary states as a result that IceCat-2 'improves the angular uncertainty areas by a factor∼5'. Since the catalog has already been processed, the results should be stated in the same tense and modality in both places.","section":"Abstract and Summary"},{"comment":"The phrase '340 events out of the original 3481' is confusing because the footnote marker makes it read as the number 3481. Please format it as '348' with the footnote, and state the original IceCat-1 event count explicitly in the main text so the reader does not have to infer it from the footnote.","section":"Sec. 2, catalog contents bullet"},{"comment":"The paper does not report the numerical angular separations between the three TDE positions and the corresponding best-fit IceCat-2 directions, or how many sigma (using the new 1σ uncertainties) each separation represents; adding these numbers would make the 'well outside' claim quantifiable and easier for readers to assess.","section":"Sec. 3, TDE paragraph and Fig. 6"},{"comment":"The distributions in Fig. 3 show IceCat-1 and IceCat-2 containment areas as histograms, but the text does not give the median IceCat-1 values from which the factor-of-5/4 improvement is computed; reporting them (e.g., in the caption or text) would let readers verify the ratio directly.","section":"Sec. 2.1, Fig. 3"},{"comment":"Reference [8] is a GCN Circular; if a peer-reviewed description of the hybrid reconstruction algorithm exists or is in preparation, it would be helpful to cite it alongside the proceedings contributions [9–11], since the validation of the catalog's uncertainties ultimately depends on that work.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"This is a proceedings-style manuscript (ICRC2025) presenting preliminary results, which lowers the bar for some presentation issues. However, if the paper is intended for a journal, the missing coverage validation and the over-strong TDE conclusion are load-bearing and would need to be addressed before the central claims can be accepted. The RA-randomization null and the transparent IceCat-1 comparison are strengths that should be preserved. The fit to the journal's scope is appropriate for astro-ph.HE / instrumentation; the main question is whether the validation can be added within the scope of a revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: IceCat-2 is exactly what it says on the tin—a reprocessed, extended version of the public IceCube alert-track catalog with a new hybrid reconstruction that shrinks the reported localization regions. The paper is a proceedings write-up, not a full methods paper, and it should be read as a resource announcement plus a set of re-examined associations. The strongest part is the catalog work itself: 365 events, consistent selection, IceTop veto applied retroactively, and a clean cross-correlation test against five source catalogs using an RA-randomization null. The observed counts match chance expectation, which is a useful sanity check and consistent with the smaller contours.\n\nWhat's genuinely new: the updated event list, the factor ~5 reduction in median 50% containment area (0.4 sq deg) and ~4 at 90% (1.3 sq deg), and the re-evaluation of known associations. TXS 0506+056 remains inside the 90% contour for IC-170922A (though no longer inside 50%), and NGC 7469 stays inside both alerts' contours. Those are useful updates for the multimessenger community.\n\nThe soft spots are real but not disqualifying. The claim that the three TDEs are 'effectively ruling out an association' is the weakest sentence in the paper. The analysis only checks spatial containment; it does not fold in the temporal coincidence (neutrinos ~100 days after peak) or the original trials factor from the TDE searches. Yes, the TDE positions lie outside the new 90% contours, but 'ruling out' needs a joint statistical treatment, and the paper doesn't provide one. That wording should be toned down.\n\nThe second concern is validation. The containment contours are taken from the reconstruction papers (refs [9–11]); there's no coverage check on the exact 365-event sample after the IceTop veto and Gold/Bronze selection. If the 90% contours under-cover for some topologies—soft events, near-horizon tracks—the improvement factors could be optimistic, and the TDE exclusion would be unsafe. This is a fair referee request, not a fatal flaw. For a proceedings paper, deferring to a dedicated methods paper is normal, but a coverage plot for this specific sample would settle it.\n\nThe circularity burden is low, as you said. The angular uncertainties come from a calibrated reconstruction, not from fitting the associations, and the RA randomization is an independent null. No invented entities, no free parameters.\n\nOverall: this is a useful, honest, preliminary catalog paper. It deserves peer review—it will be heavily used. I'd ask the authors to (1) release the catalog or at least the event list used here, (2) include a coverage check on this sample, and (3) soften the TDE 'ruling out' to 'inconsistent with' pending a joint space-time analysis.","headline":"Useful catalog update with genuinely improved localization, but the TDE 'ruling out' is overclaimed and the containment calibration deserves a coverage check before the catalog is widely used.","tokens_in":11317,"tokens_out":2374,"would_cite":true,"duration_ms":25524,"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":"Reprocessing IceCube's alert tracks with a new reconstruction framework shrinks the median 50% and 90% containment areas to 0.4 and 1.3 square degrees, and places three proposed tidal disruption event counterparts outside the updated 90%…","keywords":["IceCube","neutrino astronomy","multimessenger astronomy","track-like alerts","angular resolution","event catalog","tidal disruption events","real-time alerts"],"falsifier":"Simulate neutrino-induced muon tracks with known true directions across the energies and topologies present in the catalog, apply the new reconstruction and its containment calculation, and measure the fraction of trials in which the true direction lands inside the 50% and 90% contours; if that fraction falls short of 50% and 90% for any class of events, the factor-of-five area improvement and the TDE exclusion would be called into question.","tokens_in":10332,"feed_emoji":"🧊","tokens_out":7206,"duration_ms":69735,"temperature":0.7,"pith_summary":"This paper presents IceCat-2, a preliminary update to IceCube's public catalog of high-energy muon neutrino track alerts, spanning May 2011 to January 2025 with 365 events. Its central claim is that reprocessing every alert with the reconstruction strategy introduced in September 2024 shrinks the median 50% and 90% containment areas to 0.4 and 1.3 square degrees, a factor of about 5 and 4 below IceCat-1. The updated positions change which candidate sources remain plausible: the blazar TXS 0506+056 still sits inside the 90% region for IC-170922A but no longer the 50% one, NGC 7469 stays inside both relevant contours, and the tidal disruption events AT2019dsg, AT2019fdr, and AT2019aalc now fall well outside the updated 90% regions, effectively ending those associations. A catalog-wide search for gamma-ray and X-ray sources inside the new contours finds only as many coincidences as chance would predict. The paper argues that the improved localization makes multimessenger follow-up searches substantially more efficient and changes the list of astrophysical sources worth chasing.","feed_headline":"Fivefold sharper IceCube alert maps rule out TDE matches","feed_subtitle":"Reprocessing 365 neutrino tracks with sharper reconstructions moves tidal disruption events outside the 90% error boxes.","key_machinery":"The load-bearing mechanism is the updated track reconstruction framework adopted by IceCube's real-time alert system in September 2024, which applies different reconstruction algorithms depending on the event's energy deposition and topology. This hybrid approach improves angular resolution across a broad energy range and provides robust statistical coverage of the reported localization uncertainties, with the technical details delegated to earlier papers. The machinery produces the smaller containment contours and the revised best-fit directions that drive every result in the catalog.","core_discovery":"On its own terms, the paper's discovery is that the hybrid reconstruction framework rolled out by IceCube in September 2024, which selects different algorithms according to each event's energy deposition and topology, markedly improves the angular localization of muon track alerts while keeping the uncertainty contours statistically calibrated. For the 340 IceCat-1 events that survive the updated vetos, 50% of the new directions move by less than 0.55 degrees and 90% by less than 1.62 degrees relative to IceCat-1. The median 50% and 90% containment areas are 0.4 and 1.3 square degrees, roughly five and four times smaller than before, and the spread of areas across events narrows by a factor of 7 to 9. Re-examining previously discussed source coincidences with these sharper contours, the paper finds the TDE candidates are outside the new 90% regions, while TXS 0506+056 and NGC 7469 remain spatially compatible. Cross-checks against gamma-ray and X-ray catalogs yield coincidence counts consistent with random expectation.","pith_inferences":["If the containment calibration holds, stacking and population searches that use these alert directions should gain roughly the factor-of-five sensitivity implied by the smaller areas, enough to strengthen or decisively dismiss marginal source-class hints.","Because half the events move by more than 0.5 degrees on the sky, archival multimessenger searches that used IceCat-1 positions should be re-run with IceCat-2 directions rather than simply scaled to smaller error boxes.","An independent calibration check on simulations with known true directions, or on a known celestial shadow, would confirm whether the new 50% and 90% contours cover the true neutrino direction at the claimed rates for all alert topologies."],"forward_implications":["Follow-up observations for future alerts will search a much smaller patch of sky, lowering the cost of finding electromagnetic counterparts.","TXS 0506+056 remains inside the 90% containment region for IC-170922A but no longer inside the 50% region, so the spatial match is weakened but not removed.","NGC 7469 stays inside the updated contours for both IC-220424A and IC-230416A, so that proposed association survives the reprocessing.","The tidal disruption event associations AT2019dsg, AT2019fdr, and AT2019aalc now fall outside the updated 90% containment regions, effectively ruling them out as counterparts to the corresponding alerts.","Across the 365 alerts, the number of 90% contours containing catalog sources matches the number expected from random coincidences for 4FGL-DR4, 3FHL, 3HWC, TeVCat, and Swift-BAT, and the drop in expected coincidences relative to IceCat-1 tracks the improved resolution."],"supporting_citations":[{"why":"Defines IceCat-1, the earlier catalog whose events and containment areas IceCat-2 reprocesses and compares against.","marker":"[7]"},{"why":"Announces the September 2024 reconstruction strategy that motivates the catalog update.","marker":"[8]"},{"why":"Provide the technical description of the updated reconstruction algorithms and their treatment of systematic uncertainties.","marker":"[9–11]"},{"why":"Reports the TXS 0506+056 association with IC-170922A that IceCat-2 re-evaluates with new contours.","marker":"[4]"},{"why":"Proposes NGC 7469 as a possible neutrino emitter from two alert coincidences that IceCat-2 re-tests.","marker":"[12]"},{"why":"Propose the tidal disruption event associations AT2019dsg, AT2019fdr, and AT2019aalc that IceCat-2 places outside the updated regions.","marker":"[13–15]"},{"why":"Describes the real-time alert system whose track-like events form the catalog sample.","marker":"[3]"}],"fun_headline_variants":["IceCube's 5x sharper tracks eject TDE candidates from maps","Reprocessed neutrino alerts shrink to 0.4 sq deg, no TDEs","New IceCube catalog: 5x better resolution kills TDE links","IceCat-2 rerun with improved fits removes TDE coincidences","Sharper IceCube alerts: TDE associations now excluded"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The analysis assumes that the 50% and 90% containment contours produced by the new reconstruction are accurately calibrated for every alert, and the paper cites earlier IceCube papers rather than validating this here, so if coverage is overestimated for some event topologies, the conclusions drawn from those contours would not be safe.","fun_headline_variants_meta":{"raw":{"variants":["IceCube's 5x sharper tracks eject TDE candidates from maps","Reprocessed neutrino alerts shrink to 0.4 sq deg, no TDEs","New IceCube catalog: 5x better resolution kills TDE links","IceCat-2 rerun with improved fits removes TDE coincidences","Sharper IceCube alerts: TDE associations now excluded"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001314,"raw_usage":{"total_tokens":5380,"prompt_tokens":999,"completion_tokens":4381,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":615,"completion_tokens_details":{"reasoning_tokens":4283}},"tokens_in":615,"tokens_out":4381,"duration_ms":27645,"temperature":1.0,"reasoning_tokens":4283,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T19:08:35.709843+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Simulate neutrino-induced muon tracks with known true directions across the energies and topologies present in the catalog, apply the new reconstruction and its containment calculation, and measure the fraction of trials in which the true direction lands inside the 50% and 90% contours; if that fraction falls short of 50% and 90% for any class of events, the factor-of-five area improvement and the TDE exclusion would be called into question.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Announces the September 2024 reconstruction strategy that motivates the catalog update."}],"review_version":1}