{"id":"ea3f9be4-42b1-4b7e-969c-4a6cb5187a86","arxiv_id":"2607.18408","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"Keck imaging resolves the lens and source of OGLE-2014-BLG-0676 and, when folded into the light-curve fit, gives a 0.60-solar-mass host at 1.88 kpc with a 3.11-Jupiter-mass planet.","lead":"New Keck images taken 6.3 years after a microlensing event resolve the foreground lens star and background source, exposing a 0.5-magnitude discrepancy with earlier light-curve models. Adding the imaging data to the model shrinks the uncertainty on the host star's mass and distance, yielding a 3-Jupiter-mass planet around a 0.6-solar-mass star.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Keck χ² term is 27–34 for 4 constraints (~5σ), so the claimed 'consistent' solution is actually in significant tension with the imaging data.","rationale":"The reader's weakest assumption focused on the main-sequence mass-luminosity relation and the D_S prior calibration. I agree those are legitimate secondary concerns, but the more immediate and quantitative problem is that the best-fit model's χ²_Keck is 27–34 for four constraints. This is a statistically significant inconsistency (p < 10^-5) that the authors do not acknowledge. If the Keck data are truly incompatible with the light curve at this level, the derived physical parameters are the result of a compromise fit, not a robust measurement. The D_S prior issue would affect uncertainty estimates, but the χ²_Keck issue casts doubt on the central claim itself. My verdict remains CONDITIONAL/UNCHANGED because the paper could potentially resolve this by showing that the Keck uncertainties were underestimated (e.g., by including PSF systematics) and re-evaluating the credible intervals, but the current version does not establish the claimed robustness. I therefore recommend the same conditional outcome, with an additional required revision: explicitly report and interpret the χ²_Keck p-value and per-constraint pulls.","tokens_in":18843,"tokens_out":5803,"duration_ms":50898,"concrete_test":"Re-run the MCMC and, at the best-fit parameters, compute the posterior-predictive p-value for χ²_Keck under 4 dof, and the individual standardized residuals (pulls) for K_S, K_L, μ_rel,HN, μ_rel,HE. If the p-value is <0.01 or any pull exceeds ~3σ, the model is not statistically consistent with the imaging data, and the quoted uncertainties are unreliable. Report these pulls explicitly; this will show whether the tension is driven by one observable (e.g., K_S) or is global.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that adding Keck constraints yields a robust, self-consistent mass/distance measurement. But Table 4 lists χ²_Keck = 27.4–34.0 for the four Keck measurements (K_S, K_L, μ_rel,HN, μ_rel,HE). For 4 dof, χ²≥27 has p≈1.2×10^-5 (χ²=34: p≈7×10^-7), i.e., a 4.5–5σ joint inconsistency. The model is not simultaneously compatible with the light curve and the imaging, even though χ²_lc/dof≈1.001; it achieves a compromise rather than a self-consistent fit. This directly undermines the headline result: M_host, D_L, and m_p are derived from the intersection of these inconsistent constraints and may be biased. The paper does not discuss the magnitude or significance of χ²_Keck. The tension could stem from underestimated Keck errors, the main-sequence mass-luminosity assumption, extinction values, or source/lens misidentification — any of which would shift the mass-distance curves in Figures 2 and 5. The D_S-prior double-counting concern raised by the reader is secondary; even if it were fixed, the χ²_Keck inconsistency would remain.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reanalyzes the planetary microlensing event OGLE-2014-BLG-0676 by combining Keck/OSIRIS adaptive-optics imaging obtained 6.3 years after the event with MOA, OGLE, and Wise light-curve modeling. The authors resolve the lens and source, measure their K-band magnitudes and relative proper motion, and add these measurements as chi-square constraints in an MCMC light-curve fit. Their adopted solution gives a host mass M_host = 0.60^{+0.17}_{-0.14} Msun, lens distance D_L = 1.88^{+0.63}_{-0.35} kpc, and planet mass m_p = 3.11^{+1.11}_{-0.63} MJ, and they argue that the previous longer-timescale solution was biased. They also predict separation and magnitudes for future Roman observations.","tokens_in":19139,"tokens_out":5299,"duration_ms":47397,"significance":"If the measurement is correct, the paper is a valuable demonstration of how high-resolution imaging can break light-curve degeneracies in planetary microlensing events. The direct resolution of lens and source, the careful PSF modeling, the jackknife treatment of frame-to-frame systematics, and the concrete Roman predictions are genuine strengths. The central claim, however, is not supported by the model's own reported chi-square values: the Keck constraints are jointly inconsistent with the light-curve model at the 4-5 sigma level, and the prior on the source distance appears to have been calibrated using the same Keck data that later enter the likelihood. The headline masses and distances therefore cannot be considered robust at the claimed confidence without a substantial revision of the treatment of the imaging constraints.","major_comments":[{"comment":"The reported chi-square values for the Keck constraints are chi2_Keck = 27.4-34.0 for four measurements (K_S, K_L, mu_rel,HN, mu_rel,HE). With 4 degrees of freedom, these correspond to p ~ 1e-5 to 1e-7 (about 4-5 sigma), meaning the light-curve model and the Keck imaging are not simultaneously consistent. The paper does not discuss this tension, and the 'd.o.f.' column in Table 4 includes only the light-curve data points, not the Keck terms. This directly undermines the claim in Sections 6.2-6.3 that the solution is 'self-consistent' and 'physically reliable'. The authors should report the p-value of chi2_Keck, identify which constraint(s) dominate the discrepancy, and either propagate the tension into the parameter uncertainties or correct the systematic error budget in the Keck measurements.","section":"Table 4, Eq. (11)-(12)"},{"comment":"The prior on the source distance is described as 'constructed using a Galactic model implemented via genulens, and calibrated to be consistent with the observational constraints from the Keck image analysis, as described in Section 3.' If this means the D_S prior was adjusted to match the same Keck measurements that are later used in chi2_Keck, then the imaging information is used twice, biasing the posterior and artificially narrowing the quoted uncertainties. This is a load-bearing circularity. The authors must either demonstrate that the D_S prior was derived from the Galactic model alone, or re-run the fit with a prior that is not informed by the Keck data.","section":"Section 4.2.1"},{"comment":"The derived lens mass and distance depend on converting the measured K-band lens brightness to a mass through a main-sequence mass-luminosity relation, and on the adopted extinction values A_K,rc = 0.37 and A_I,rc = 2.50. The paper acknowledges that a white-dwarf lens cannot be entirely excluded, but it does not propagate this as an alternative model. Given that the Keck constraints are already in tension with the light curve, the mass-luminosity and extinction assumptions are plausible sources of the discrepancy. The authors should provide a sensitivity analysis that allows the lens to be a white dwarf (or an unresolved binary) and that varies the extinction parameters over their full ranges, and show how the resulting M_L-D_L posteriors change.","section":"Sections 3 and 4.2.2, Eqs. (3)-(5), (8)-(9)"}],"minor_comments":[{"comment":"The photometric transformations in Eqs. (6)-(7) need a clearer statement of the photometric systems and the covariance among the coefficients. The presence of the -1.00+/-0.29 mag offset in Eq. (7) is large and should be justified explicitly.","section":"Equations (6)-(7)"},{"comment":"In Section 3 the lens brightness is quoted as K_L = 16.98 +/- 0.05 mag, while Table 2 lists K_L = 16.982 +/- 0.080 mag. The uncertainty differs; please reconcile the two values.","section":"Section 3 and Table 2"},{"comment":"The 'MCMC Averages' column appears to contain two entries for several parameters (e.g., s and alpha) within a single cell. This makes the table difficult to read; separate the close and wide values into distinct rows or columns.","section":"Table 4"},{"comment":"The text says the adopted pixel scale is 10 mas/pixel while the measured scale is 9.952 mas/pixel, and that 'the difference does not affect our result.' Given that the measured separations are at the few-mas level, a brief quantitative statement of the resulting shift would be useful.","section":"Section 2.1"}],"recommendation":"major_revision","confidential_remarks":"The chi2_Keck values in Table 4 are the most direct problem: the paper's central claim of a consistent joint solution is contradicted by its own numbers. The D_S prior circularity is also concerning. I would be willing to reconsider after the authors address these two issues, but the current manuscript is not acceptable as is."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know this one has a load-bearing flaw hiding in plain sight. The paper resolves the lens and source with Keck AO, measures the proper motion and both fluxes, and re-fits the light curve with those as constraints. That is the right program, and the observational work is careful: two-star PSF fitting, jackknife systematic errors, a clean identification of which component is which. The new measurement of the source being 0.5 mag brighter than the old model prediction is real, and the authors are right that something had to give.\n\nBut look at Table 4. The four Keck constraints contribute χ²_Keck = 27–34 across the four best-fit models. For four independent measurements, that is p ~ 10⁻⁵ to 10⁻⁷ — a 4.5–5σ mismatch between the model and the imaging data. The paper never reports this number or its significance. Instead it presents the combined fit as self-consistent, and the quoted mass, distance, and planet mass are exactly the quantities most sensitive to this tension. The fit is a compromise, not a solution.\n\nThe reader’s circularity concern about the D_S prior is secondary but real. Section 4.2.1 says the prior is “calibrated to be consistent with the observational constraints from the Keck image analysis” and then those same constraints enter as likelihood terms. That double-counts the imaging data and likely overstates precision. Even if you fixed that, though, the χ²_Keck inconsistency remains.\n\nWhat the paper does well: the photometric reduction is serious, the systematic error treatment is above average, the Roman predictions are concrete, and the identification of the brighter component as the lens is reasonable though not ironclad. The method itself is established — this is an application, not a new technique — but for this specific event it produces a genuinely new measurement and a clear challenge to the previous t_E solution.\n\nThe fix is straightforward in principle. The authors need to report χ²_Keck and its p-value, investigate whether the Keck error bars are underestimated (they should check frame-to-frame systematics more carefully and consider covariance between the two proper-motion components), and re-run with a D_S prior built from the Galactic model alone, not calibrated to the Keck data. If the tension survives, the conclusion changes. If it dissolves with renormalized errors, the paper is publishable.\n\nI’d send this back for major revision. The observational result is worth refereeing carefully, but the central claim is not established in the current version. Serious referee time is warranted; the paper is coherent and honest on its own terms, just missing the most important diagnostic.","headline":"The Keck-constrained fit is not actually consistent with the Keck data — χ²_Keck ≈ 27–34 for 4 constraints is a ~5σ problem the paper never acknowledges.","tokens_in":19745,"tokens_out":2554,"would_cite":false,"duration_ms":25828,"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":"High-resolution imaging of the microlensing event OGLE-2014-BLG-0676 breaks the light-curve degeneracy and shows the system is a 3.11-Jupiter-mass planet around a 0.60-solar-mass star at 1.88 kpc.","keywords":["gravitational microlensing","exoplanets","adaptive optics","high-angular-resolution imaging","light-curve modeling","OGLE-2014-BLG-0676","mass-luminosity relation","degeneracy breaking"],"falsifier":"A future high-resolution observation of this field that measures the lens and source in multiple near-infrared bands: if the lens appears more than ~1 magnitude fainter than predicted for a 0.6-solar-mass main-sequence star at 1.88 kpc (K_L ≈ 16.7), or if the source flux is inconsistent with the model's prediction, the claimed mass and distance would be refuted.","tokens_in":18683,"feed_emoji":"🪐","tokens_out":6846,"duration_ms":51474,"temperature":0.7,"pith_summary":"The paper claims that follow-up high-resolution imaging of a microlensing event can do more than refine parameters—it can expose a biased light-curve model and correct it. For the planetary event OGLE-2014-BLG-0676, adaptive-optics images taken 6.3 years after the event resolved the lens and source separately. The measured source brightness disagreed with the previously published model, and the paper shows that the earlier long-timescale solution was an artifact of degeneracies. Re-fitting the light curve with the imaging constraints incorporated yields a self-consistent solution: a 3.11-Jupiter-mass planet orbiting a 0.60-solar-mass M/K dwarf at about 1.9 kpc. The result matters because it demonstrates a general strategy for breaking microlensing degeneracies without relying on higher-order effects.","feed_headline":"Keck snapshots pin a microlensing planet at 3 Jupiter masses","feed_subtitle":"Adaptive-optics imaging resolves lens and source, giving host mass 0.6 solar masses at 1.9 kpc.","key_machinery":"The central mechanism is image-constrained light-curve modeling: a combined chi-square (χ²_total = χ²_lc + χ²_Keck) that penalizes light-curve models whose predicted source flux, lens flux, and geocentric relative proper motion disagree with the measured resolved-imaging values. The lens-flux prediction uses an empirical main-sequence mass–luminosity relation, while the proper-motion constraint converts the measured heliocentric separation into the geocentric frame used by the light-curve model. This combined penalty breaks the Einstein-timescale/source-flux degeneracy that left the earlier model biased.","core_discovery":"Incorporating the resolved-lens measurements—the lens's K-band magnitude, the source's K-band magnitude, and the lens–source relative proper motion—directly into the microlensing light-curve modeling removes the degeneracy that plagued earlier fits. The imaging data show the K-band source flux is 0.52 magnitudes brighter than predicted by the previous model and the relative proper motion is about 6 mas/yr, larger than the roughly 4 mas/yr assumed. With these constraints added as a penalty term to the fit, the best-fit Einstein timescale shortens from about 100–130 days to about 90 days, and the resulting physical parameters are a host mass of 0.60+0.17−0.14 solar masses, a lens distance of 1","pith_inferences":["The same approach can be applied to other microlensing events with archival or future high-resolution follow-up imaging, potentially revising published planet masses where light-curve-only models are degenerate.","The quoted masses are conditional on the resolved brighter star being a single main-sequence lens; if it is a white dwarf or unresolved binary, the mass–distance relation shifts and the host mass and distance would change.","A second epoch of high-resolution imaging would directly test the predicted relative proper motion and check for a bound companion, the main remaining alternative interpretation.","The 0.52-magnitude source-flux discrepancy implies that other parameters derived from the same light-curve data set (e.g., the angular Einstein radius) may also carry hidden biases that resolved imaging can expose."],"forward_implications":["The system is a 3.11-Jupiter-mass planet around a 0.60-solar-mass M/K dwarf, a benchmark for planet formation around low-mass stars.","The earlier 100–130 day timescale solution is disfavored; the true timescale is about 90 days, showing that sparsely covered light-curve wings can bias t_E when blending is unconstrained.","Resolving the lens and source directly eliminates the need to assume that unresolved blended light originates from the lens, a simplification earlier work was forced to make.","The image-constrained fits shrink the uncertainties on host mass, planet mass, distance, and separation compared with Bayesian-only or unresolved-imaging analyses.","Future survey observations of this field can confirm the predicted lens and source magnitudes and separation, and refine the parameters with multi-band photometry."],"fun_headline_variants":["Keck AO resolves lens and source, pinning planet at 3 MJup","Microlensing planet mass nailed by resolved lens imaging","AO imaging ends degeneracy, gives precise host and planet masses","3-Jupiter-mass planet confirmed via Keck adaptive optics","Resolved lens breaks degeneracy, pins mass of OGLE-2014-BLG-0676L"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The central claim rests on the assumption that the brighter resolved component is the lens and that it is a single main-sequence star following the adopted mass–luminosity relation; if the lens is a white dwarf, an unresolved binary, or the adopted extinction values are incorrect, the derived host mass and distance would not hold.","fun_headline_variants_meta":{"raw":{"variants":["Keck AO resolves lens and source, pinning planet at 3 MJup","Microlensing planet mass nailed by resolved lens imaging","AO imaging ends degeneracy, gives precise host and planet masses","3-Jupiter-mass planet confirmed via Keck adaptive optics","Resolved lens breaks degeneracy, pins mass of OGLE-2014-BLG-0676L"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00042,"raw_usage":{"total_tokens":2063,"prompt_tokens":874,"completion_tokens":1189,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":618,"completion_tokens_details":{"reasoning_tokens":1093}},"tokens_in":618,"tokens_out":1189,"duration_ms":9871,"temperature":1.0,"reasoning_tokens":1093,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T15:29:26.547065+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A future high-resolution observation of this field that measures the lens and source in multiple near-infrared bands: if the lens appears more than ~1 magnitude fainter than predicted for a 0.6-solar-mass main-sequence star at 1.88 kpc (K_L ≈ 16.7), or if the source flux is inconsistent with the model's prediction, the claimed mass and distance would be refuted.","supporting_citations":[],"review_version":1}