{"id":"80ada4d0-4ac0-47de-aac9-33a6766a5d05","arxiv_id":"2411.16197","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Giant planets around M dwarfs show lower bulk metallicities than those around FGK stars, with similar mass-metallicity slopes but a significant offset between the populations.","lead":"Using thermal evolution models, the authors inferred the heavy-element content of giant exoplanets and found that those orbiting M-dwarf stars appear metal-poor compared to those around sun-like FGK stars. The result, if correct, points to different formation conditions around the galaxy's most common stars.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The M-dwarf metallicity offset in Eq. 3 may be inflated by the homogeneous-adiabatic interior assumption: Section 4 itself notes that slow growth around M dwarfs is expected to cause non-adiabatic cooling and composition gradients, which would bias inferred metallicities low.","rationale":"I agree with the reader's weakest_assumption. The strongest claim is explicitly a statement about inferred bulk metallicities, and the inference depends on a structural assumption that the authors themselves identify as likely violated for M dwarfs. Unlike other caveats, this one has a clear direction: it would raise the true M-dwarf metallicities relative to the reported values, shrinking the offset. The paper's Section 4 defense that model uncertainties apply equally to both populations is not responsive to this effect, because the formation timescale difference is the very thing that makes the interiors non-adiabatic for M dwarfs. A quantitative test exists and is feasible. The paper's other limitations (20 planets, age-prior sensitivity, mass-range selection) are real but either acknowledged and bounded in Appendix D or would not by themselves erase the offset; the interior-model bias is the single unresolved threat. Therefore the reader's CONDITIONAL verdict is appropriate, and I would not change it.","tokens_in":22827,"tokens_out":5759,"duration_ms":186696,"concrete_test":"Re-run the mass-limited M-dwarf retrievals (0.3–2 M_J, 15 planets) with non-adiabatic, composition-gradient evolution models (e.g., Vazan & Helled 2020) or with an equivalent thermal-state correction, keeping the same observed mass, radius, and age priors. Refit Eq. 3 and compare the M-dwarf intercept and its posterior with the FGK intercept. If the M-dwarf intercept moves upward by more than roughly one sigma of the current difference (i.e., toward the FGK value) or the two intercepts become consistent within 2 sigma, the claimed population offset is not robust to interior modeling.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim rests on planetsynth's homogeneous, adiabatic thermal evolution models to convert measured masses and radii into bulk metallicities. The paper's Section 4 explicitly states that slow growth around M dwarfs 'is expected to lead to very extended composition gradients and non-adiabatic cooling for a large portion of the interior,' making real M-dwarf interiors hotter than the homogeneous models. For a fixed observed radius, a hotter interior can hold more heavy elements, so the inferred Z and Z/Z* for M-dwarf giants is biased low. This bias is not covered by the paper's assertion that model uncertainties affect both populations equally, because it is specific to the M-dwarf formation regime. Since the claimed offset in Eq. 3 is driven by the M-dwarf intercept (1.71 ± 0.73 versus 8.01 ± 1.19), an unquantified low-Z bias of this type could convert a real population difference into an artifact. The age-prior and small-sample issues are secondary; this is the one assumption whose failure would most directly erase the headline result.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This Letter by Müller and Helled uses the planetsynth thermal evolution models (built on MESA) to infer bulk metallicities and heavy-element masses for warm giant planets with reliable mass–radius measurements, split into planets around FGK stars (104 successful retrievals) and around M dwarfs (20 successful retrievals). For a mass-limited sample of 0.3–2 Jupiter masses, Bayesian regression yields power-law relations in Eq. (3): Z/Z* = (8.01±1.19) M^(-0.41±0.17) for FGK hosts and Z/Z* = (1.71±0.73) M^(-0.74±0.60) for M dwarfs, with the intercepts stated to be inconsistent by more than 2σ. The paper concludes that the currently available data suggest a lack of metal-rich giant planets around M dwarfs compared to FGK-star counterparts, and it also reports a moderate correlation between stellar metallicity and planetary heavy-element mass for FGK hosts. The appendix tests the influence of age priors on the M-dwarf results and examines various correlations.","tokens_in":23250,"tokens_out":5129,"duration_ms":48878,"significance":"If the central claim holds, the result is an important observational constraint on giant-planet formation across stellar mass, suggesting that the heavy-element content of giant planets around M dwarfs is systematically lower than around FGK stars, possibly due to different formation conditions such as lower disk solid content or a different accretion history. The authors provide a transparent presentation of their sample, the adopted priors, and the posterior distributions of all fit parameters, and they explicitly test the sensitivity to age priors in Appendix D. The use of a published, MESA-validated evolution model and a robust Bayesian regression with Student-t likelihoods are strengths of the analysis. The significance of the result is, however, limited by the small M-dwarf sample and by the dependence of the inferred metallicities on the homogeneous-adiabatic interior assumption, which the authors themselves identify as a potential source of bias for slowly formed M-dwarf planets.","major_comments":[{"comment":"The paper argues, in the second paragraph of Section 4, that model uncertainties (equation of state, atmospheric model, interior profiles) apply equally to planets around FGK and M-dwarf stars, so that relative differences between the populations are preserved. This defense is incomplete because the final paragraph of the same section states that slow growth around M dwarfs is expected to lead to very extended composition gradients and non-adiabatic cooling for a large portion of the interior, making real interiors hotter than the homogeneous adiabatic models. For a fixed observed radius, a hotter planet can contain more heavy elements, so the inferred Z and Z/Z* for M-dwarf giants would be biased low, and this bias is specific to the M-dwarf formation regime. Since the claimed offset in Eq. (3) is driven by the M-dwarf intercept (1.71±0.73 versus 8.01±1.19), an unquantified low-Z bias of this type could in principle erase the headline result. Please provide a quantitative estimate of this effect using non-homogeneous or composition-gradient evolution models, or explicitly reframe the conclusion as conditional on the homogeneous-adiabatic assumption.","section":"Section 4"},{"comment":"Appendix D demonstrates that if the eight M dwarfs without stellar age estimates were all very young (1–3 Gyr), the M-Mz difference between the populations would drop from about 2σ to about 1σ. Because the nominal age prior of 1–10 Gyr for these eight objects is arbitrary, and because Section 3.2 states that the metal-poor result for M dwarfs is \"robust,\" the robustness claim is too strong without quantifying the prior probability of the young-age scenario or extending the age-prior test to the Z/Z* quantity used in the headline relation. Please report the significance of the Z/Z* offset under both extreme age priors and discuss the implications of the ~1σ outcome for the central claim.","section":"Appendix D"},{"comment":"The choice of the mass-limited sample, 0.3 ≤ Mp(MJ) ≤ 2, is partly motivated by \"the masses of the current observed giant planets around M-dwarf stars,\" as stated in Section 3.1. This makes the mass window of the central comparison selected on the very M-dwarf sample whose metallicity is being characterised, which is a source of potential selection bias. The full-sample fits in Appendix B behave differently (e.g., the negative M-Mz slope for M dwarfs), and the authors themselves prefer the mass-limited sample partly for that reason. Because the headline offset in Eq. (3) depends on this post-hoc mass cut, please justify the 0.3–2 MJ range from independent grounds (e.g., from the FGK population alone or from formation theory) and show how the inferred offsets change under reasonable variations of the upper cut (e.g., 1.5, 2.5, and 3 MJ).","section":"Section 3.1"}],"minor_comments":[{"comment":"The description of the prior distribution on the fit parameters, \"p(β0, β1) ∝ N(µ = 0, σ = 1)\", is imprecise; it should be written as independent normal priors on β0 and β1, each with mean 0 and standard deviation 1.","section":"Section 2"},{"comment":"The statement that the two populations have \"similar slopes\" for the M-Z/Z* relations refers only to the mass-limited sample; for the full sample in Appendix B, Eq. (B.1) gives slopes that are formally inconsistent (−0.71 ± 0.04 versus −1.42 ± 0.19). Please make this qualification explicit in the abstract and conclusions.","section":"Abstract"},{"comment":"Appendix D would be clearer if the text stated which 8 of the 20 M-dwarf planets have unknown ages and which 12 have stellar-age estimates, so that the reader can evaluate the representativeness of the re-analysis without cross-referencing Table 1.","section":"Appendix D"},{"comment":"Several rows in Table 1 have missing entries for age or metallicity; please add a footnote clarifying that missing values indicate either no successful interior retrieval or no published age/metallicity estimate, to avoid ambiguity.","section":"Table 1"}],"recommendation":"major_revision","confidential_remarks":"The paper is likely suitable for A&A Letters if the authors can address the three load-bearing concerns: (i) the direction and magnitude of the composition-gradient/non-adiabatic bias for M-dwarf planets, (ii) the sensitivity of the headline offset to extreme age priors shown in their own Appendix D, and (iii) the selection of the mass window based partly on the M-dwarf sample. The central claim is interesting and the analysis is generally careful, but the robustness statements currently outrun the evidence. In addition, I would encourage the editor to consider whether the journal's word limit permits the requested quantitative checks; if not, the authors may need to shorten other parts of the manuscript or move material to an online appendix."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Simon, here's my read on Müller & Helled (arXiv:2411.16197).\n\nThe genuinely new thing is the population comparison: applying established thermal-evolution retrieval to a sample of giant planets around M dwarfs and asking whether their mass–metallicity relation differs from FGK hosts. That hasn't been done before, and the result — a lower intercept for M dwarfs (Z/Z* around 1.7 vs 8 at 1 MJ) — is worth taking seriously. The paper does a lot right: the sample is carefully built from the PlanetS catalog with recent M-dwarf discoveries added, the models are validated against MESA, the EOS dependence is discussed, and the appendices give posteriors, sensitivity tests, and correlation checks. The authors are also unusually candid about their limitations.\n\nThe soft spots are real, though. The M-dwarf sample is 20 planets (15 in the preferred mass range), and the 0.3–2 MJ window was chosen partly because that's where the M dwarfs are, which makes the 'population' comparison less clean. Appendix D shows that if the eight M dwarfs with unknown ages are actually young (1–3 Gyr), the offset drops to roughly 1 sigma. That's the biggest statistical vulnerability.\n\nThe modeling vulnerability is the one you flagged. The paper's Section 4 explicitly says that slow growth around M dwarfs should produce extended composition gradients and non-adiabatic cooling, so the real interiors are likely hotter than the homogeneous-adiabatic models assume. For a fixed radius, a hotter planet can hold more heavy elements, so the inferred M-dwarf Z would be biased low — exactly in the direction of the claimed offset. The authors' argument that model uncertainties affect both populations equally doesn't cover this, because this effect is specific to the M-dwarf formation regime. They acknowledge it but don't quantify it. The age-prior bias works the other way (their generous 1–10 Gyr prior biases M-dwarf Z high), so the net sign is unclear, but that uncertainty alone means the headline offset isn't yet robust.\n\nNet: this is a solid, transparent paper that deserves a serious referee. The result is plausible and important if true, but the current evidence is suggestive, not conclusive. A referee should ask for a quantitative estimate of the non-adiabatic/gradient bias on the M-dwarf sample, or at least a statement that the offset could be partly an artifact. I'd take it to peer review and see it through revision.","headline":"First population-level claim of lower giant-planet bulk metallicity around M dwarfs, but the offset rests on a small sample and an unquantified interior-model bias that runs in the direction of the result.","tokens_in":23660,"tokens_out":3253,"would_cite":true,"duration_ms":71491,"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 claims that giant planets orbiting M-dwarf stars contain systematically less heavy material than giant planets around FGK stars, with the two populations' mass–metallicity trends separated mainly by a lower offset for M-dwarf…","keywords":["giant planets","M dwarfs","bulk metallicity","mass–metallicity relation","thermal evolution models","planet interiors","exoplanet formation","exoplanet characterization"],"falsifier":"Re-derive the M-dwarf metallicity offsets with non-adiabatic, compositionally stratified evolution models on the same planets; if those models raise the inferred $Z/Z_*$ values into agreement with the FGK relation, the claimed population difference is an artifact of the homogeneous, adiabatic assumption. Alternatively, measure atmospheric metallicities for several M-dwarf giants in the 0.3–2 $M_J$ range; atmospheric values systematically above the bulk inferences would signal the same bias.","tokens_in":22609,"feed_emoji":"🪐","tokens_out":10670,"duration_ms":89774,"temperature":0.7,"pith_summary":"This paper asks whether giant planets that orbit cool M-dwarf stars are built from less heavy material than giant planets around Sun-like stars. Using thermal evolution models to convert 20 M-dwarf and 104 FGK-host mass–radius measurements into bulk heavy-element fractions, the authors find that both populations follow inverse mass–metallicity relations with statistically similar slopes, but the M-dwarf relation sits at a lower offset: for masses from 0.3 to 2 Jupiter masses, the normalized metallicity at a given mass is about five times smaller than for FGK stars. They conclude that current data lack the metal-rich giant planets around M dwarfs, and they interpret this deficit as a signature of different formation conditions around low-mass stars. The result matters because it connects planetary composition to the mass of the host star, giving formation and evolution models a new population-level constraint.","feed_headline":"M-dwarf giants hold fewer heavy elements than FGK giants","feed_subtitle":"Their mass–metallicity offset from FGK-host planets exceeds two sigma, hinting at different formation paths.","key_machinery":"The central object is planetsynth, a grid of thermal-evolution models computed with a stellar-evolution code modified for giant planets; the models assume homogeneous, adiabatic interiors with a hydrogen–helium equation of state. For each planet, Monte Carlo sampling draws mass, radius, and age from observed priors and finds the heavy-element fraction whose modeled cooling radius matches the measured radius, building a posterior distribution for $Z$ and heavy-element mass $M_z$. A monotonicity-preserving interpolation and extrapolation scheme extends the radius evolution past the models' original 10-gigayear limit. The mass–metallicity and mass–heavy-element relations come from Bayesian linear regression in log–log space with a Student-$t$ likelihood and Hamiltonian Monte Carlo sampling. This machinery is what converts an observed mass–radius point into a position on the $Z$–$M$ plane where the two host-star populations can be compared.","core_discovery":"On the paper's own terms, the discovery is a population offset: giant planets around M dwarfs are systematically metal-poor compared with giant planets around FGK stars. In the preferred mass-limited sample ($0.3\\le M_p/M_J \\le 2$), the inferred normalized-metallicity relations are $Z/Z_* = (8.01\\pm1.19) M^{-0.41\\pm0.17}$ for FGK hosts and $Z/Z_* = (1.71\\pm0.73) M^{-0.74\\pm0.60}$ for M dwarfs, with intercepts inconsistent at more than two standard deviations; the heavy-element masses are $M_z = (46.81\\pm9.86) M^{0.37\\pm0.24}$ and $M_z = (20.17\\pm6.83) M^{0.40\\pm0.53}$ in Earth masses. The authors also report that for FGK hosts the bulk metallicity and residual heavy-element mass correlate moderately with stellar metallicity, while the small M-dwarf sample shows no significant such correlation. They attribute the lower M-dwarf offset to a lack of metal-rich giant planets, consistent with longer core-accretion timescales, fewer available solids, or formation by disk instability with little post-formation heavy-element accretion. They caution that poorly known M-dwarf ages and interior-model assumptions could shift absolute values, but argue the population difference itself is robust.","pith_inferences":["An editor's inference: the homogeneous, adiabatic interior assumption is the main uncontrolled variable; if slowly accreting M-dwarf planets really develop composition gradients and non-adiabatic cooling, the inferred bulk metallicities could shift upward enough to partially or fully close the gap.","A testable prediction follows from the astrophysical interpretation: systematic searches for giant planets around M dwarfs should find a deficit of high-heavy-element-mass planets at fixed $M_p$, which upcoming transit surveys can check.","Transmission-spectrum atmospheric metallicities of the M-dwarf giants provide an independent probe; atmospheres that are richer in heavy elements than the inferred bulk would point to the interior-model bias rather than to a truly metal-poor population.","The FGK stellar-metallicity correlation found here, if it persists in larger samples, gives formation models a quantitative target that earlier smaller studies could not resolve."],"forward_implications":["Both populations show statistically compatible mass–metallicity slopes in the 0.3–2 $M_J$ range, so the difference between them is a vertical offset rather than a different scaling with mass.","If the offset is real, giant-planet formation around M dwarfs must deliver systematically less heavy material, favoring formation paths that either grow cores slowly in low-mass disks or assemble by disk instability with limited post-formation solid accretion.","The traditional picture of a fixed critical core mass of about 10 $M_\\oplus$ for runaway gas accretion becomes questionable around M dwarfs; formation models should test whether gas accretion can begin with significantly smaller cores.","More transit and radial-velocity detections of 0.3–2 $M_J$ planets around M dwarfs directly test the result by shrinking the uncertainty on the M-dwarf intercept.","Atmospheric metallicity measurements of M-dwarf giants would connect their atmospheres to their inferred bulk interiors and help break the degeneracy that interior characterization alone cannot resolve."],"supporting_citations":[{"why":"Supplies the updated exoplanet catalog whose mass, radius, age, and irradiation measurements define most of the sample.","marker":"Parc et al. (2024)"},{"why":"Provides the planetsynth thermal-evolution model grid used to infer each planet's heavy-element fraction.","marker":"Müller & Helled (2021)"},{"why":"Underlying stellar-evolution code, modified for giant planets, that computes the cooling models.","marker":"Paxton et al. (2011, 2013, 2015, 2018, 2019)"},{"why":"Current release of the evolution code used in the model grid.","marker":"Jermyn et al. (2023)"},{"why":"Establishes the Monte Carlo approach for converting mass–radius measurements into bulk compositions.","marker":"Miller & Fortney (2011)"},{"why":"Previous mass–metallicity relation that this work compares against and updates.","marker":"Thorngren et al. (2016)"},{"why":"Hydrogen–helium equation of state whose higher density lowers inferred metallicities relative to earlier equations of state.","marker":"Chabrier et al. (2019)"},{"why":"Basis for the expectation that slowly growing M-dwarf planets have composition gradients and non-adiabatic cooling, the central modeling caveat.","marker":"Helled & Stevenson (2017)"},{"why":"Adds recently detected M-dwarf giant planets to the sample.","marker":"Hartman et al. (2024)"}],"fun_headline_variants":["M-dwarf giants are metal-poor vs FGK giants","Giant planets around M stars lack heavy elements","M-star giants show lower bulk metallicity","Metal-poor giants orbit M dwarfs more often"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The result assumes that a measured radius maps to a unique heavy-element fraction through standard well-mixed, evenly cooling interior models; if slowly formed M-dwarf giants really have layered compositions and cool unevenly, their interiors would be hotter than modeled and their inferred metallicities could be biased low, possibly erasing the reported population offset.","fun_headline_variants_meta":{"raw":{"variants":["M-dwarf giants are metal-poor vs FGK giants","Giant planets around M stars lack heavy elements","M-star giants show lower bulk metallicity","Metal-poor giants orbit M dwarfs more often"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000168,"raw_usage":{"total_tokens":1342,"prompt_tokens":1109,"completion_tokens":233,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":725,"completion_tokens_details":{"reasoning_tokens":171}},"tokens_in":725,"tokens_out":233,"duration_ms":2901,"temperature":1.0,"reasoning_tokens":171,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T13:22:58.415596+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-derive the M-dwarf metallicity offsets with non-adiabatic, compositionally stratified evolution models on the same planets; if those models raise the inferred $Z/Z_*$ values into agreement with the FGK relation, the claimed population difference is an artifact of the homogeneous, adiabatic assumption. Alternatively, measure atmospheric metallicities for several M-dwarf giants in the 0.3–2 $M_J$ range; atmospheric values systematically above the bulk inferences would signal the same bias.","supporting_citations":[{"cited_title":"2019, , 872, 51","cited_arxiv_id":null,"evidence_quote":"Hydrogen–helium equation of state whose higher density lowers inferred metallicities relative to earlier equations of state."},{"cited_title":"& Stevenson , D","cited_arxiv_id":null,"evidence_quote":"Basis for the expectation that slowly growing M-dwarf planets have composition gradients and non-adiabatic cooling, the central modeling caveat."},{"cited_title":"TOI 762 A b and TIC 46432937 b: Two Giant Planets Transiting M Dwarf Stars","cited_arxiv_id":"2407.07187","evidence_quote":"Adds recently detected M-dwarf giant planets to the sample."}],"review_version":1}