{"id":"5eec992b-2722-46f0-aaec-0e8be343b834","arxiv_id":"2508.08364","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"The wide-binary fraction of white dwarfs declines steeply with white dwarf mass and their eccentricities are abnormally low, modeled as evidence that mass-loss timescales shorten sharply for high-mass stars.","lead":"White dwarfs in wide binaries are rarer and less eccentric than expected from their main-sequence predecessors. The authors use Gaia data plus dynamical models to argue that slow mass loss stretches orbits for low-mass stars, while fast mass loss in high-mass stars disrupts them, giving new constraints on a poorly observed phase of stellar evolution.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Inference of short τ_AGB for massive stars is degenerate with the unmeasured initial eccentricity distribution of high-mass MS binaries; the model's quantitative failure to reproduce the observed decline adds to the concern.","rationale":"The reader's weakest_assumption identifies exactly the concern I find most load-bearing: the unknown initial eccentricity distribution of high-mass MS binaries. The paper's own Sec 4.4 flags this as a limitation, and the reader correctly notes that it could change the inferred τ_AGB(mass). My stress-test agrees with this assessment. I considered whether the more fundamental issue is the model's quantitative failure (factor ~2 vs. >6 decline), but that failure is acknowledged in the paper and does not by itself falsify the claim that a mass-dependent τ_AGB is required — the model still produces a decline only when τ_AGB is mass-dependent. The initial eccentricity distribution, however, could potentially reproduce the observed decline and the eccentricity split without a short τ_AGB, directly undermining the central 'requires' statement. Therefore, this is the single most load-bearing concern. The proposed concrete test is computational and would settle whether the degeneracy actually breaks the claim. Since the concern reinforces the reader's CONDITIONAL verdict rather than overturning it, the recommended verdict remains UNCHANGED.","tokens_in":25608,"tokens_out":7392,"duration_ms":87947,"concrete_test":"Using the published code (github.com/zakamska/binaries), rerun the fiducial population synthesis model with the initial eccentricity power-law index α for high-mass progenitors (initial mass > 2 M_sun) varied over a grid (e.g., α_high = -0.5, 0, 0.5, 1.0, 1.5, 2.0), while keeping the α(a) relation for low-mass stars fixed. For each α_high, test whether any constant τ_AGB (e.g., 10^5 yr, 10^6 yr) can simultaneously reproduce (i) the observed WD-MS binary fraction decline in Fig. 1 and (ii) the observed eccentricity split in Fig. 9 within the quoted 68% credible intervals. If such a combination exists, the claim that a mass-dependent τ_AGB with short timescales is required would be falsified; if none exists, the inference survives this test.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that the steep decline in WD-MS binary fraction requires a mass-dependent τ_AGB (10^3–10^4 yr for high-mass stars) rests on the population synthesis model's assumed initial eccentricity distribution for high-mass MS binaries: p(e) ∝ e^α with α(a) from Hwang et al. (2022b), which is calibrated on solar-type MS binaries and extrapolated to masses where it is not measured. Sec 4.4 explicitly states this distribution 'is not well known for high-mass MS stars.' This matters because both observables used to infer τ_AGB(mass) — the binary-fraction decline (Fig. 1) and the eccentricity split between massive and low-mass WD-MS binaries (Fig. 9) — are sensitive to the initial eccentricities. Higher birth eccentricities for massive progenitors would increase disruption during mass loss and recoil, steepening the binary-fraction decline even with a constant τ_AGB, and would raise the final eccentricities of surviving massive WD-MS binaries, mimicking the signature currently attributed to short τ_AGB (Sec 4.4). The inference is therefore degenerate. This concern is compounded by the admitted quantitative mismatch: the fiducial mass-dependent τ_AGB model reproduces only a factor ~2 decline while the data show a factor >6 (Sec 5). While the direction of the trend may still favor some mass dependence, the specific 10^3–10^4 yr bound is not robust until the initial eccentricity distribution for high-mass stars is independently constrained or the model is tested against it.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper combines Gaia wide-binary catalogs with population synthesis to study how post-main-sequence mass loss and recoil shape the orbits of wide binaries containing white dwarfs. Observationally, it reports that the WD-MS wide-binary fraction (10^3–10^4 AU) is roughly constant at low WD mass and then declines steeply by a factor ~6 above 0.6 Msun, that WD-WD binary fractions mirror this decline, and that WD-MS and WD-WD binaries have lower eccentricities than MS-MS binaries at the same separations. The modeling part computes orbital evolution for mass-loss timescales spanning the secular to impulsive regimes, with a recoil tied to mass loss through momentum conservation (Eq. 4). The authors find that adiabatic expansion explains the low eccentricities, while reproducing the steep mass-dependent binary fraction requires a mass-dependent mass-loss timescale, declining from 2×10^6 yr at 1 Msun to 10^3 yr at 6 Msun. They interpret the data as requiring short (10^3–10^4 yr) mass-loss timescales for high-mass stars, and they confirm earlier recoil constraints of ~0.25–4 km/s.","tokens_in":25973,"tokens_out":4517,"duration_ms":55060,"significance":"If the central inference holds, the paper provides a novel dynamical constraint on the poorly observed post-AGB mass-loss phase, and it does so with an unusually careful treatment of timescales from adiabatic to impulsive regimes. The observational sample is thoughtfully selected, with explicit completeness and robustness checks (parallax cuts, ruwe, photometry), and the paper is transparent about its model limitations, including the unmeasured initial eccentricity distribution of high-mass binaries and the quantitative mismatch with the observed factor >6 decline. A significant strength is the public availability of the population-synthesis code (GitHub/Zenodo), which makes the modeling reproducible. The predicted eccentricity difference between massive and low-mass WD-MS binaries (Fig. 9) is a falsifiable signature that could be tested with larger samples or independent high-mass eccentricity constraints. The paper should be considered a valuable contribution if the main degeneracy is addressed and the central claim is appropriately re-scaled to what the models actually demonstrate.","major_comments":[{"comment":"The inference of short tau_AGB for high-mass stars is degenerate with the unmeasured initial eccentricity distribution of high-mass MS binaries. The paper states this explicitly: the initial eccentricity distribution from Hwang et al. (2022b) 'is not well known for high-mass MS stars.' Because both the binary-fraction decline (Fig. 8) and the eccentricity split (Fig. 9) are generated from p(e) ∝ e^α with α(a) calibrated on solar-type binaries, a higher birth α for massive progenitors would steepen the disruption-driven decline and raise the final eccentricities of survivors, mimicking the signature attributed to short tau_AGB. This is a load-bearing degeneracy for the central claim. Please quantify it by repeating the population synthesis with α(a) varied over plausible ranges for high-mass stars, or by using independent eccentricity constraints for B-type / massive MS binaries. Without","section":"Sec. 4.4 and Sec. 3.2"},{"comment":"The quantitative mismatch between the fiducial model and the data is admitted: the model produces only a factor ~2 decline in WD-MS, WD-WD, and retention fractions, while the observed decline is a factor >6. This limits the strength of the conclusion. The statement that the steep decline 'requires' short tau_AGB should be softened to something like 'is directionally consistent with' or 'is necessary but not sufficient' until the model can reproduce the observed amplitude. Moreover, the tau_AGB=10^4 yr model is said to 'perform similarly' to the fiducial 10^3 yr model, so the upper bound rests on a factor-2 reproduction of a much steeper trend. Please provide a quantitative goodness-of-fit or likelihood criterion for what counts as reproducing the mass dependence, and discuss whether the gap can be closed by plausible ingredients discussed qualitatively in Sec. 4.5 (mass-dependent triple","section":"Sec. 5 and Fig. 8"},{"comment":"The mass-dependent tau_AGB is a free parameter chosen to fit the observed mass dependence of the binary fraction, and then the same model is used to interpret the eccentricity data. This is a partial circularity: the eccentricity split in Fig. 9 is not an independent confirmation if the tau_AGB(m) relation was already tuned to the mass-dependent binary fraction. To strengthen the predictive claim, fix tau_AGB(m) independently from stellar evolution models (e.g., Miller Bertolami 2016, with an explicit uncertainty band) and test whether the eccentricity predictions still match; alternatively, report a joint fit statistic over binary fraction and eccentricity. As written, the eccentricity comparison is a consistency check, not a confirmation.","section":"Sec. 4.3–4.4"}],"minor_comments":[{"comment":"Typo: 'dusty think disk' should be 'dusty thin disk'.","section":"Sec. 2.1"},{"comment":"The text refers to 'Cumming et al. (2008)' for the initial-to-final mass relation, but the intended reference appears to be Cummings et al. (2018). Please correct the citation.","section":"Sec. 2.4"},{"comment":"Equation (6) is derived for a power-law distribution on e ∈ [0,1], but the text says alpha is computed over truncated ranges e ∈ [0,e_max] and then the median over several e_max is reported. Please state explicitly which likelihood is maximized when e_max < 1, since Eq. (6) does not apply verbatim in that case.","section":"Eq. (5)-(6) and Sec. 3.2"},{"comment":"The paper says 'we ignore the differences between intrinsic semi-major axes and observed separations for the purposes of this calculation.' Given that the eccentricity comparison uses observed projected separations, please quantify the impact of projection effects on the inferred alpha, or justify why they are subdominant.","section":"Sec. 3.2"},{"comment":"The left panel would benefit from an explicit note that the horizontal dashed line is the predicted v-r angle distribution for a thermal eccentricity distribution after folding at 90 deg; the current caption is slightly terse.","section":"Fig. 5"}],"recommendation":"major_revision","confidential_remarks":"The paper is honest and the code availability is a real strength. However, the central astrophysical claim—short tau_AGB for massive stars—rests on a degeneracy with the initial eccentricity distribution of high-mass binaries that the authors themselves acknowledge, and the fiducial model reproduces only a factor ~2 of the observed factor >6 decline. These are not fatal flaws, but they require additional modeling and a more careful statement of what is actually proven. I recommend major revision rather than rejection, because the observational data and the dynamical framework are valuable and the inference can be sharpened within the paper's scope."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThis one is worth your time for the measurements. The Gaia-based wide-binary fractions for WD-MS and WD-WD systems as a function of WD mass, and the eccentricity distributions from v-r angles, are genuinely new, carefully selected, and come with sensible robustness checks (parallax cuts, ruwe, crowding). The modeling is also a step forward: they handle mass loss across the full timescale range, couple the kick to the mass-loss rate, and show adiabatic expansion plausibly explains why WD binaries have lower eccentricities than MS-MS ones at the same separation. They also confirm the El-Badry & Rix recoil constraint with a different model setup.\n\nThe soft spot is the central quantitative claim. The fiducial mass-dependent tau_AGB model only reproduces a factor ~2 decline in the binary fraction where the data show a factor >6; the authors say this themselves in Sec 5. The eccentricity split between high- and low-mass WD-MS binaries is suggestive but sits inside large error bars. And the stress-test concern is legitimate: the initial eccentricity distribution for high-mass MS stars is unmeasured, and both observables used to infer short tau_AGB for massive stars depend on it. If massive progenitors are born on higher-eccentricity orbits, disruption and final eccentricities both change in the direction that mimics short tau_AGB. The paper flags this in Sec 4.4 but does not quantify it. So the headline 'requires' a 10^3-10^4 yr timescale is stronger than the evidence supports. What the data robustly show is that the binary fraction declines steeply with WD mass; the mass-dependent tau_AGB is a plausible explanation, but not uniquely forced.\n\nOther minor concerns: the MS-MS baseline is stitched across surveys with a manual normalization at 1 Msun, adding systematic uncertainty to the retention fraction; and the binary fraction definition, though self-consistent, has narrow companion selection, so the numbers should be compared with care.\n\nCredit where due: the paper is honest about its limitations, the code and data are public, and the writing is clear.\n\nThis is for stellar evolution and wide-binary audiences. I'd cite it for the observations. It deserves a serious referee; the right referee should press on the eccentricity degeneracy and ask for a quantitative model comparison rather than a qualitative one.\n\nReading group: maybe, if people care about white dwarf evolution or binary dynamics.","headline":"Strong new Gaia measurements of WD wide binary fractions and eccentricities; the dynamical claim of a mass-dependent mass-loss timescale is plausible but not quantitatively nailed, and the short-timescale inference is partly degenerate with unknown birth eccentricities of massive stars.","tokens_in":26460,"tokens_out":2479,"would_cite":true,"duration_ms":29135,"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":"Wide binaries containing white dwarfs reveal that the final mass-loss phase of massive stars lasts about a thousand years—a thousand times shorter than for Sun-like stars.","keywords":["binary stars","white dwarfs","wide binaries","stellar mass loss","Gaia","eccentricity distribution","population synthesis","late stellar evolution"],"falsifier":"Measure the eccentricities of high-mass ($m_{\\rm WD} > 0.8\\,M_\\odot$) versus low-mass ($0.5$–$0.8\\,M_\\odot$) WD-MS binaries at $1000$–$3000$ AU with a larger sample. The model predicts $\\alpha \\approx 1.2$–$1.4$ for massive WD-MS binaries if the mass-loss timescale is $\\sim 10^3$ yr, versus $\\alpha \\approx 0.6$–$0.8$ if it is $\\sim 10^5$ yr, so a clean measurement of the split would confirm or reject the fast mass-loss scenario.","tokens_in":25473,"feed_emoji":"💫","tokens_out":16212,"duration_ms":161590,"temperature":0.7,"pith_summary":"The paper claims that wide binaries—pairs of stars bound at separations from a few hundred to tens of thousands of times the Earth–Sun distance—record the final mass-loss phase of stellar evolution, and that Gaia's census of white dwarf companions allows that phase to be timed. The observations show two opposing trends: white dwarfs become much less likely to have a wide main-sequence companion as their mass rises above $0.6\\,M_\\odot$, even though main-sequence stars become more likely to have wide companions with mass; and white dwarf binaries are significantly less eccentric than main-sequence binaries at the same separation. A population model including slow (adiabatic) and fast (impulsive) mass loss, plus velocity recoil tied to the mass loss, reproduces both trends only when the mass-loss timescale drops sharply with mass—from roughly two million years at $1\\,M_\\odot$ to a thousand to ten thousand years at $6\\,M_\\odot$. Slow mass loss inflates orbits and explains the low eccentricities; fast mass loss in massive stars disrupts their wide companions and explains the falling binary fraction. This gives a dynamical, observation-based handle on a stellar phase that is hard to see directly.","feed_headline":"Heavy stars shed their final mass in about 1,000 years","feed_subtitle":"Gaia white-dwarf binaries tie stellar mass loss to a sharp timescale drop from ~2 million years to ~1,000 years.","key_machinery":"The machinery is the dynamical response of a binary to mass loss and recoil across three timescale regimes, plus one mass-dependent curve that picks the regime. When the mass-loss timescale $\\tau_{\\rm AGB}$ greatly exceeds the orbital period, the Delaunay action $L$—an adiabatic invariant—is conserved and the orbit expands, $a \\propto 1/(m_1+m_2)$; when $\\tau_{\\rm AGB}$ is much shorter, mass is lost impulsively and the widest orbits are disrupted. Systems in between are integrated numerically. Recoil is tied to mass loss by momentum conservation, $d v_k/dt = v_{\\rm asym}\\,(\\dot m_2/m_2)$, and in the slow regime acts as a Stark-like secular perturbation on eccentricity. The pivotal adjustable","core_discovery":"Two Gaia findings anchor the paper: the fraction of white dwarfs with a main-sequence wide companion is flat near 3% below $0.6\\,M_\\odot$ and falls about sixfold by $1.2\\,M_\\odot$, while white-dwarf–containing binaries are markedly less eccentric than MS-MS binaries at the same separations. The model's key move is to let mass loss act on any timescale relative to the orbit. Slow adiabatic loss inflates orbits, $a \\propto 1/(m_1+m_2)$, pulling tighter, rounder binaries into the observed separation range and explaining the low eccentricities. The steep decline then forces a mass-loss timescale that drops with mass—from $\\sim 2\\times10^6$ yr at $1\\,M_\\odot$ to $\\sim 10^3$ yr at $6\\,M_\\odot$ in","pith_inferences":["If the mass-dependent $\\tau_{\\rm AGB}$ is a real clock, then wide-binary statistics at fixed white dwarf mass can be used to map how the AGB/post-AGB transition depends on metallicity and age—something the single-population model here does not split.","The same adiabatic-versus-impulsive framework could be applied to wide binaries containing neutron stars or black holes, potentially constraining their kicks and the timescale of core-collapse mass loss, once such samples become large enough.","The model's factor-of-two decline in the fiducial case falls short of the observed factor-of-six; a fully self-consistent treatment of hierarchical triples, which the paper only discusses qualitatively, may close the gap or force an even steeper $\\tau_{\\rm AGB}(m)$.","Future Gaia data releases with more massive MS-MS binaries could measure the birth eccentricity distribution at the high-mass end directly, turning the eccentricity split in Figure 9 into a clean empirical determination of the mass-loss duration for ~3–4 $M_\\odot$ stars."],"forward_implications":["Massive stars (≳3 $M_\\odot$) must shed their envelopes on ≲10^4-year timescales; stellar evolution models that keep such stars in a slow-loss regime will fail to explain Gaia's white dwarf binary statistics.","The low eccentricities of WD-MS and WD-WD binaries are a signature that their orbits expanded adiabatically: the binaries observed at 10^3–10^4 AU were born at smaller separations, so eccentricity measurements become a tracer of orbital expansion.","Recoil velocities of order 0.25–1 km/s, tied to mass loss via momentum conservation, are required to steepen the wide-separation distribution; zero-recoil models are excluded.","White dwarfs below 0.5 $M_\\odot$ in wide binaries are predominantly the surviving bright components of close binaries in hierarchical triples, so their statistics open a separate window into common-envelope evolution and triple dynamics.","High-mass white dwarfs should retain fewer exo-Oort comets and show less metal pollution than low-mass ones, a testable prediction for white dwarf atmospheric abundances."],"supporting_citations":[{"why":"Established the kick/disruption framework for white-dwarf wide binaries and the recoil scale; this paper builds on it and confirms its conclusion.","marker":"El-Badry & Rix (2018)"},{"why":"Supplies the v–r angle method and the measured separation-dependent eccentricity distribution p(e)∝e^α used both as initial conditions and as the comparison baseline.","marker":"Hwang et al. (2022b)"},{"why":"Provides the initial-to-final mass relation that sets how much mass each progenitor loses on the way to becoming a white dwarf.","marker":"Cummings et al. (2018)"},{"why":"Theoretical post-AGB timescales that decline steeply with mass, the stellar-evolution anchor for the required mass-dependent τ_AGB.","marker":"Miller Bertolami (2016)"},{"why":"AGB models showing low-mass stars finish near their core masses while more massive stars retain envelopes to lose later, motivating a mass-dependent mass-loss phase.","marker":"Weiss & Ferguson (2009)"},{"why":"Supplies the steeply rising main-sequence wide-binary fraction with mass used to define the retention fraction and to contrast with the white-dwarf decline.","marker":"Moe & Di Stefano (2017)"},{"why":"The white-dwarf catalog and membership probabilities from which the target sample is selected.","marker":"Gentile Fusillo et al. (2021)"},{"why":"The Gaia wide-binary catalog and comoving selection used to identify WD-MS and WD-WD pairs.","marker":"El-Badry et al. (2021)"}],"fun_headline_variants":["Heavy stars shed last mass in ~1,000 years","Mass loss timescale collapses for massive stars","Gaia: heavy stars lose final mass 2000x faster","Why heavy stars lose mass fast: white-dwarf binaries","Stellar mass loss: heavy stars finish in a millennium"],"cache_read_input_tokens":2816,"weakest_assumption_plain":"The model assumes that wide binaries that will become massive white-dwarf systems are born with the same separation-dependent eccentricity distribution measured for lower-mass main-sequence binaries, but that birth distribution is not well known for high-mass stars; if it differs, the inferred mass-loss timescale changes.","fun_headline_variants_meta":{"raw":{"variants":["Heavy stars shed last mass in ~1,000 years","Mass loss timescale collapses for massive stars","Gaia: heavy stars lose final mass 2000x faster","Why heavy stars lose mass fast: white-dwarf binaries","Stellar mass loss: heavy stars finish in a millennium"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000294,"raw_usage":{"total_tokens":1616,"prompt_tokens":884,"completion_tokens":732,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":628,"completion_tokens_details":{"reasoning_tokens":651}},"tokens_in":628,"tokens_out":732,"duration_ms":8682,"temperature":1.0,"reasoning_tokens":651,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T21:31:35.469172+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the eccentricities of high-mass ($m_{\\rm WD} > 0.8\\,M_\\odot$) versus low-mass ($0.5$–$0.8\\,M_\\odot$) WD-MS binaries at $1000$–$3000$ AU with a larger sample. The model predicts $\\alpha \\approx 1.2$–$1.4$ for massive WD-MS binaries if the mass-loss timescale is $\\sim 10^3$ yr, versus $\\alpha \\approx 0.6$–$0.8$ if it is $\\sim 10^5$ yr, so a clean measurement of the split would confirm or reject the fast mass-loss scenario.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"AGB models showing low-mass stars finish near their core masses while more massive stars retain envelopes to lose later, motivating a mass-dependent mass-loss phase."}],"review_version":1}