{"id":"7b5c049f-9f94-4219-afc2-fedb0a087bd6","arxiv_id":"2603.20371","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"New radial-velocity orbits refute nearly every high-mass dark-companion candidate in Gaia DR3; only Gaia BH1 and BH2 remain as black holes, with one new 1.16 Msun compact-object candidate in the acceleration sample.","lead":"This paper checked dozens of black-hole and neutron-star candidates identified in the Gaia DR3 binary catalog with new telescope measurements. It finds that nearly all of the high-mass candidates are false, leaving the two previously confirmed black holes, and points to acceleration-catalog stars as the next place to look.","discovery_kind":"replication","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Census claim that only two BH binaries exist in Gaia DR3 depends on an unvalidated completeness assumption for low-mass-function and massive-primary systems.","rationale":"The reader's CONDITIONAL verdict already captures the main risk, so no verdict change is needed. I partially agree with the reader's weakest_assumption: sparse or phase-incomplete RV orbits are a real concern for individual objects, but for most systems the paper's falsification of the Gaia orbit does not require a fully determined alternative orbit — a large mismatch between observed and predicted RV amplitude is sufficient. The more load-bearing issue for the paper's central census claim is sample completeness. The direct catalog cuts exclude low-mass-function systems and systems where the luminous primary is more massive than the dark companion; the 2–5 Msun range is covered only through external candidate lists. The paper explicitly assumes no face-on orbits (footnote 9), and for SB1-only systems this assumption is not checked. This does not overturn the paper's demonstration that the selected Gaia orbits are wrong, but it does mean the categorical conclusion 'only two wide binaries with black hole secondaries are contained in the Gaia DR3 binary catalogs' is not fully established by the presented data. The proposed catalog re-query is a concrete way to test completeness, and DR4 astrometry will provide an independent check.","tokens_in":30944,"tokens_out":15706,"duration_ms":167104,"concrete_test":"Re-run the Section 2.1 selection on the full Gaia DR3 non-single-star tables without the m2_lower > 5 and m2_lower > m1 cuts, and without relying on Shahaf/Andrews lists: select all P > 9 d orbital solutions with no SB2/eclipsing/UV-bright flags, recompute companion-mass posteriors from the published orbital elements and the paper's stellar parameters, and list any system with M2 > 2 Msun absent from Table 1. If the list is non-empty, the 'only two BH binaries' census is incomplete. As a complementary check, for each SB1-only source rejected in Section 3.2, compute the inclination that would make M2 > 2 Msun and test it against Gaia RUWE/epoch astrometry or Gaia DR4.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The per-object refutations of the selected Gaia orbits are mostly strong: direct RV measurements contradict the predicted large-amplitude orbits, and forward modeling with gaiamock supports several exclusions. The load-bearing weakness is the stronger census statement in the abstract and conclusions: 'only two wide binaries with black hole secondaries are contained in the Gaia DR3 binary catalogs.' Section 2.1.1 selects directly only sources with m2_lower > 5 Msun and m2_lower > m1; for the 2–5 Msun interval the paper relies entirely on the Shahaf et al. and Andrews et al. candidate lists. A true black hole binary in the DR3 SB1 or astrometric catalogs with low inclination (small RV mass function) or with a primary more massive than the secondary would not enter the sample. Footnote 9 explicitly concedes the face-on assumption, and for SB1-only candidates this is not individually tested with astrometry/RUWE. The paper's own text for Gaia DR3 5846362195472084992 — 'we cannot currently rule out a higher mass secondary' — illustrates that even within the selected sample, the absence of a massive companion is not always conclusively established. Thus the data demonstrate that the particular Gaia orbits examined are wrong, but they do not establish that no other DR3 orbital solution hides a massive dark companion.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper presents spectroscopic radial-velocity follow-up of 31 systems from Gaia DR3: 20 stars with orbital solutions (astrometric, SB1, or AstroSpectroSB1) flagged as candidate black-hole/neutron-star binaries and 11 stars with large accelerations but no orbital solution. The authors derive stellar parameters and new RV orbits using MIKE, APF, IMACS, FEROS, and HIRES data, supplemented by archival velocities. Their main result is that, apart from the previously confirmed Gaia BH1, Gaia BH2, and Gaia NS1, the Gaia DR3 orbital solutions for all other selected candidate systems with inferred dark companions above 2 Msun are inconsistent with the new RVs; most are lower-amplitude binaries with low-mass companions, several are SB2 or eclipsing systems, and a few hot stars show no detectable variability. They propose stricter quality cuts (significance > 10, F2 < 4) and report one acceleration source with Msini = 1.16 +/- 0.01 Msun, likely an ultramassive white dwarf or neutron star. The paper concludes that only two wide binaries with black-hole secondaries are contained in the Gaia DR3 binary catalogs.","tokens_in":31330,"tokens_out":8525,"duration_ms":90472,"significance":"The per-object refutations are largely convincing: the RV measurements are independent of the Gaia values being tested, the paper is transparent about ambiguous cases, and forward modeling with gaiamock and the Lam et al. framework strengthens the interpretations for several systems. If the per-object results stand, the paper substantially sharpens our understanding of the false-positive rate in the high-mass-function tail of the Gaia DR3 binary catalogs and gives practical guidance for DR4. The identification of the acceleration source with Msini = 1.16 Msun is also a valuable result. However, the global census conclusion - 'only two wide binaries with black hole secondaries are contained in the Gaia DR3 binary catalogs' - is a load-bearing claim that goes beyond the selected sample and needs to be either relaxed or supported by a completeness argument.","major_comments":[{"comment":"The abstract and conclusions state that only two wide binaries with black-hole secondaries are contained in the Gaia DR3 binary catalogs. This census claim is not supported by the sample definition. In Sec. 2.1.1 the direct catalog selection requires m2_lower >= 5 Msun and m2_lower > m1; candidates with inferred dark companions between 2 and 5 Msun are included only if they appear in the Shahaf et al. or Andrews et al. lists. A genuine black-hole binary viewed at low inclination (small RV mass function) or with a primary more massive than the secondary would not enter the sample. Footnote 9 explicitly concedes the face-on assumption for the accelerating sources, but the same limitation applies to the SB1 and astrometric candidates. The data demonstrate that the particular Gaia orbits examined are wrong, but they do not establish that no other DR3 orbital solution hides a massive dark com","section":"Abstract; Sec. 5; Sec. 2.1.1"},{"comment":"For some objects the new RV solutions are too ambiguous to support the blanket statement in Sec. 5 that all remaining candidates have brown-dwarf or M-dwarf companions. For Gaia DR3 5846362195472084992, Sec. 3.2 states that the RVs cover only half of the orbit and 'we cannot currently rule out a higher mass secondary'; Table 4 gives K1 = 12.8 +21.2/-5.6 km/s and f = 0.045 +0.183/-0.032 Msun, so the allowed companion mass extends well above 1 Msun. Similarly, for Gaia DR3 3640889032890567040, Sec. 3.1 presents two acceptable RV solutions, with the longer-period solution giving Msini = 1.30 Msun. Both systems are inconsistent with their claimed Gaia orbits, so the core refutation stands, but the per-object caveats should be carried into the conclusions rather than replaced by a stronger universal statement.","section":"Sec. 3.2; Table 4; Sec. 5"},{"comment":"The proposed quality cuts (significance > 10, F2 < 4) are fit to the same sample they are used to clean. The thresholds are chosen post hoc to retain the three known compact-object binaries and exclude the false solutions in Table 1; no independent validation set is used, and the paper does not quantify how many genuine binaries would be lost. The statement that the cuts 'would retain the known genuine compact objects while excluding all but one of the false solutions' is a description of the training sample, not a validated selection function. This is a recommendation rather than the paper's central proof, but it should be explicitly labeled provisional, and ideally tested on simulated Gaia-like catalogs or withheld DR4-like data before being applied to future catalogs.","section":"Sec. 4.1"}],"minor_comments":[{"comment":"The header abstract says 'radial velocity orbital solutions for 12 binaries' while the body abstract and Table 4 list 11 solutions. Please make the number consistent.","section":"Abstract"},{"comment":"As printed, Table 2 contains entries for only one star. Please state explicitly that the full machine-readable table is available and describe its format.","section":"Table 2"},{"comment":"The hot-star metallicity prior (-0.2 <= [Fe/H] <= 0.2) is introduced to avoid unphysical metallicities. Its effect on the inferred primary masses and hence on the companion-mass limits should be quantified or at least discussed.","section":"Sec. 2.5.2"},{"comment":"For Gaia DR3 3509370326763016704, the interpretation that the astrometric signal could be due to an almost-resolved luminous tertiary is plausible but speculative; the sentence 'something beyond a brown dwarf orbiting a low-mass star may be present' could be worded more cautiously.","section":"Sec. 3.1"}],"recommendation":"major_revision","confidential_remarks":"The per-object RV results are the main strength of this paper and are, in my reading, largely sound. The issue that needs to be resolved before publication is framing: the 'only two black-hole binaries in DR3' census claim is broader than the selected sample can support. This is fixable by either softening the conclusion to the candidates actually examined or adding a rigorous completeness calculation. I would not reject the paper; the revisions can be local."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Good to talk. Quick take: the per-object refutations are the real content here, and they hold up. The census punchline — “only two wide binaries with black hole secondaries are contained in the Gaia DR3 binary catalogs” — is stronger than the data support, and I’d want that softened before it goes out. But the RV work is genuine, independent of the Gaia values, and it systematically kills the high-mass-function tail of DR3.\n\nWhat’s new: they selected a near-complete set of DR3 orbital candidates with candidate dark companions above 2 Msun (Shahaf, Andrews, plus direct cuts) and measured velocities for essentially all of them. The observed RV amplitudes are factors of several to orders of magnitude below the Gaia-predicted amplitudes for almost every object; most are low-mass companions, a few are SB2, some are hot stars with bad Gaia RVs. They also give full RV orbital solutions for 11 of them, and the one acceleration-source orbit with M2 sin i = 1.16 Msun is a nice find. The tables and figures are transparent about which constraints are secure and which aren’t.\n\nThe soft spots: First, the “only two” census conclusion rests on an untested completeness assumption. The direct catalog cut was m2_lower > 5 Msun and m2_lower > m1; the 2–5 Msun interval was covered only through the Shahaf and Andrews candidate lists, and both lists have their own selection functions. A wide BH with a low-inclination orbit (small RV mass function) or a primary more massive than the secondary would not be in this sample. Footnote 9 concedes the face-on assumption for the acceleration sources, but the paper doesn’t test it for the SB1 or astrometric candidates via RUWE or astrometry. So the data nail the specific orbits they examined; they don’t prove the catalog contains no other BH binaries.\n\nSecond, the recommended quality cuts (significance > 10, F2 < 4) are chosen post hoc to keep BH1/BH2/NS1 and drop the false positives. That’s a reasonable heuristic, but it’s not validated on an independent sample. The paper doesn’t oversell it — they say it “may be useful” — but it should be labeled as a suggestion, not a vetted prescription.\n\nThird, a few of the new RV orbits are genuinely ambiguous: 3640889 has two solutions, 5846362 covers only half the phase, and for that source they admit they can’t rule out a higher-mass secondary. These are appropriately caveated in the text, but the conclusions drift toward a confidence the per-object data don’t always support. Minor copyedit: the abstract says 12 RV solutions, the text says 11.\n\nOverall: this is a solid, useful refutation paper. It deserves a serious referee and publication after the census claim is made conditional on selection completeness. I’d bring it to the reading group — the discussion about exactly what DR3 and DR4 can deliver for BH demographics is worth hashing out.","headline":"Solid per-object RV refutation of the DR3 compact-object candidates; the 'only two BH binaries' census claim outruns the selection completeness.","tokens_in":31777,"tokens_out":3415,"would_cite":true,"duration_ms":37046,"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":"Spectroscopic follow-up shows that, apart from the three previously confirmed systems, every Gaia DR3 binary with a candidate dark companion above two solar masses has an incorrect orbital solution, leaving only two wide black-hole binaries","keywords":["Gaia DR3 binary catalog","black hole binaries","neutron star binaries","radial velocity orbits","mass function","astrometric binaries","spectroscopic binaries","compact object companions"],"falsifier":"If dense radial-velocity monitoring of Gaia DR3 3640889032890567040 over a full orbit confirms the longer-period (≈940 d) solution with a semi-amplitude near 18 km/s, that system's minimum companion mass would be 1.3 solar masses—invalidating the paper's claim that no unconfirmed DR3 candidate has a compact companion above two solar masses.","tokens_in":30907,"feed_emoji":"🔭","tokens_out":8856,"duration_ms":82966,"temperature":0.7,"pith_summary":"Gaia's third data release listed dozens of binary systems whose unseen companions were massive enough to be black holes or neutron stars. This paper takes the most promising of those candidates—single-lined binaries with inferred dark companions above two solar masses—and measures their radial velocities with ground-based spectroscopy over months to years. The result is stark: except for the three previously confirmed systems (Gaia BH1, Gaia BH2, and Gaia NS1), every one of the Gaia orbital solutions is wrong. The re-derived orbits have much smaller velocity amplitudes and different periods, putting the companions in the brown-dwarf, low-mass-star, or white-dwarf range, or revealing a second luminous star. The authors conclude that only two wide black-hole binaries exist in the entire DR3 binary catalog, and they suggest stricter quality cuts to keep future catalogs from repeating the mistake.","feed_headline":"Only two Gaia black-hole binaries survive follow-up","feed_subtitle":"New spectra show the rest have brown-dwarf, M-dwarf, or white-dwarf companions—not black holes.","key_machinery":"The binary mass function f = K1^3 P (1 - e^2)^{3/2} / (2πG), computed from the radial-velocity semi-amplitude K1, orbital period P, and eccentricity e, is the load-bearing quantity. It yields a strict lower limit on the unseen companion's mass, so a small measured K1 rules out a multi-solar-mass companion regardless of inclination. Orbital parameters are found by sampling the posterior with a rejection-sampling algorithm and refining with Markov-chain Monte Carlo, using radial velocities from new spectra plus archival data. Stellar masses of the primaries come from fitting spectra and photometry to evolutionary isochrones. The same orbital parameters are then used to predict the astrometric","core_discovery":"The authors derive independent radial-velocity orbits for eleven systems selected from the Gaia DR3 binary catalogs as having candidate dark companions above two solar masses. In every case except the previously confirmed Gaia BH1, Gaia BH2, and Gaia NS1, the spectroscopic orbit is incompatible with the published Gaia solution: periods change (for one star, from 1039 days to 20 days), and velocity semi-amplitudes are typically several times smaller than the Gaia orbit would predict. The resulting mass functions place the companions at sub-solar masses—brown dwarfs, M dwarfs, or white dwarfs—or show a second set of stellar absorption lines, indicating a luminous companion. For the one system","pith_inferences":["If the proposed quality cuts are applied retroactively to DR3, the number of 'reliable' orbital solutions with high-mass dark companions shrinks to near zero, implying that the Milky Way's wide black-hole binary population may be at the low end of pre-Gaia theoretical predictions.","The hierarchical-triple interpretation raised for at least one rejected candidate (a 20-day inner binary embedded in a roughly 3-year outer orbit) suggests that some of the discarded systems could still contain compact objects as outer components; future epoch astrometry could test this for each rejected candidate.","The accelerating system with a 1.16-solar-mass minimum companion is a natural target for Gaia DR4 astrometry, which could measure its inclination directly and settle whether the companion is a neutron star or an ultramassive white dwarf.","Combining dense radial-velocity orbits with forward models of Gaia's detection probability, as done case-by-case here, could be developed into a formal selection function for the binary catalogs, enabling unbiased population statistics instead of informal quality cuts."],"forward_implications":["The number of wide black-hole binaries known from Gaia DR3 drops from a few dozen published candidates to two, so the local census of such systems is much smaller than the raw catalog suggested.","Future Gaia catalogs should apply the stricter quality cuts suggested here—significance greater than 10 and goodness of fit below 4—to keep candidate lists clean of spurious massive companions.","Hot, rapidly rotating stars are a systematic source of bad Gaia radial velocities; candidate searches should flag them for independent confirmation rather than accepting their orbital solutions.","The acceleration catalogs, which lack full orbital solutions, are a better place than the orbital-solution catalogs to look for the next confirmed neutron-star or black-hole binary, since the one fully characterized accelerating system here is a likely neutron star or ultramassive white dwarf.","Wide black-hole binaries with periods of several years and small velocity amplitudes will be missed entirely by short monitoring campaigns, so confirming them will require sustained, multi-year spectroscopy."],"fun_headline_variants":["Most Gaia BH candidates fail follow-up spectroscopy","Gaia's black-hole finds shrink to two confirmed","Follow-up demotes most Gaia black-hole candidates","No new black holes from Gaia binary catalog"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The conclusion rests on the assumption that the new radial-velocity data—sometimes sparse or phase-incomplete—correctly identify the true orbit of each system, so that the disagreement with the Gaia solution is not hiding a hierarchical triple, an alias period, or a nearly face-on orbit with a genuinely massive companion.","fun_headline_variants_meta":{"raw":{"variants":["Most Gaia BH candidates fail follow-up spectroscopy","Gaia's black-hole finds shrink to two confirmed","Follow-up demotes most Gaia black-hole candidates","No new black holes from Gaia binary catalog"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000399,"raw_usage":{"total_tokens":1929,"prompt_tokens":759,"completion_tokens":1170,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":503,"completion_tokens_details":{"reasoning_tokens":1112}},"tokens_in":503,"tokens_out":1170,"duration_ms":9713,"temperature":1.0,"reasoning_tokens":1112,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-02T17:45:31.125015+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"If dense radial-velocity monitoring of Gaia DR3 3640889032890567040 over a full orbit confirms the longer-period (≈940 d) solution with a semi-amplitude near 18 km/s, that system's minimum companion mass would be 1.3 solar masses—invalidating the paper's claim that no unconfirmed DR3 candidate has a compact companion above two solar masses.","supporting_citations":[],"review_version":1}