{"id":"dd5e7444-a28e-4749-b3a1-8bb0e01b84fc","arxiv_id":"2501.15298","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"With Planck and DESI BAO data, the Brans-Dicke Galileon model prefers a nonzero Galileon term and yields H0 = 71.0 +1.5 -1.3 km/s/Mpc, consistent with SH0ES at 1.2 sigma.","lead":"This paper tests four scalar-tensor gravity models against the DESI 2024 BAO data combined with Planck. It finds that one model, Brans-Dicke Galileon, raises the inferred Hubble constant and shrinks the Hubble tension to 1.2 sigma.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"BDG '3σ detection' and 1.2σ H0 tension are driven by DESI's two lowest-z bins; replacing them with SDSS reduces the detection to ~2σ and raises the H0 tension to 2.6σ, so the central claim is contingent on BGS/LRG1 being unbiased.","rationale":"The paper is a competent constraints analysis that clearly states its assumptions and reports the sensitivity of its central result to the low-redshift BAO data. The reader's conditional verdict is well founded: the BDG detection and the H0-tension reduction hinge on the DESI BGS and LRG1 bins. The paper's own P18+(DESI+SDSS) comparison in Table III shows that replacing those bins lowers the Galileon parameter significance from ~3σ to ~2σ and raises the SH0ES tension from 1.2σ to 2.6σ. A failure of those bins would collapse the abstract's central claims. No internal inconsistency is apparent in the model equations or the likelihood construction, but the custom CLASSig extension is not public, so an independent numerical check would be valuable. The quoted '3σ' significance is based on a prior-truncated posterior and may be mildly overstated compared to a profile-likelihood or Δχ²-based estimate, but the dominant risk remains the low-z data dependence. For these reasons I do not see a basis to reject the paper, but the result should be treated as conditional on the robustness of the DESI low-z bins, exactly as the reader concluded. The concrete test above would settle whether the concern lands.","tokens_in":27953,"tokens_out":9978,"duration_ms":91627,"concrete_test":"Run three variants of the P18+DESI BDG analysis: (a) exclude the BGS and LRG1 bins entirely while keeping all DESI bins at z>0.8; (b) add a systematic error term to those two bins with amplitude set by the DESI systematics budget, inflating the covariance; (c) replace the two bins with the overlapping SDSS/BOSS low-z BAO points while keeping all other DESI bins unchanged. If in any of these variants the 68% credible interval for 1/eα8 includes 0 at less than 3σ, or if H0 shifts by more than 1 km/s/Mpc relative to the fiducial result, the headline detection and H0-tension reduction are not robust to plausible low-z systematics.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's headline result for BDG, 1/eα8 = 0.255+0.095−0.064 (about 3σ from the ΛCDM limit) and H0 = 71.0+1.5−1.3 km/s/Mpc (1.2σ from SH0ES), is presented for P18+DESI. The paper itself discloses that the result is 'mainly driven by the first two redshift bins of DESI' (abstract and Section IV.B). Its cross-check P18+(DESI+SDSS), which replaces those two bins (BGS and LRG1) with SDSS low-z BAO data, yields 1/eα8 = 0.157+0.088−0.076 (consistent with 0 at 2σ) and H0 = 69.2+0.9−1.2 km/s/Mpc (2.6σ tension with SH0ES), per Table III. Thus the central claim is almost entirely contingent on the DESI BGS and LRG1 measurements. If those bins suffer from systematics—for example in reconstruction, fiber assignment, or template assumptions—the claimed 3σ detection of the Galileon term and the 1.2σ H0 tension would disappear. This is not an internal inconsistency; it is an empirical fragility that the paper honestly reports but does not resolve. The fixed value of ξ = 5×10−5 and the lack of public likelihood code are secondary concerns, but the low-z bin sensitivity is the most load-bearing because it can change the qualitative conclusion.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper analyzes four scalar-tensor gravity models (induced gravity IG, induced gravity with an effective gravitational-constant imbalance DeltaIG, Brans-Dicke Galileon BDG, and early modified gravity with conformal coupling EMG-CC) against Planck 2018 CMB data combined with DESI 2024 BAO data, with variants replacing the two lowest DESI redshift bins by SDSS low-z BAO data and with SH0ES/CCHP H0 priors. The main claim is that BDG prefers a nonzero Galileon term, 1/eα8 = 0.255+0.095−0.064, about 3σ from the ΛCDM limit, and H0 = 71.0+1.5−1.3 km/s/Mpc, reducing the SH0ES tension to 1.2σ, while the same parameter is only an upper limit with previous SDSS BAO data. The paper also reports that this preference is mainly driven by the first two DESI redshift bins, and that replacing them with SDSS data lowers the Galileon significance to roughly 2σ and increases the H0 tension to 2.6σ. The IG, DeltaIG, and EMG-CC models do not show a statistical preference over ΛCDM with CMB+BAO alone.","tokens_in":28337,"tokens_out":6086,"duration_ms":55433,"significance":"If the BDG detection were robust, this would be an interesting and timely result: a concrete modified-gravity model that simultaneously explains the DESI BAO preference for dynamical dark energy and substantially alleviates the Hubble tension, with a full treatment of background and perturbations rather than a phenomenological parametrization. The authors are to be credited for performing a complete Einstein-Boltzmann treatment, for comparing four models on equal footing, and for explicitly reporting the cross-check with SDSS low-z data instead of hiding the fragility. However, the central claim is highly sensitive to the low-redshift DESI bins, and the paper contains internal inconsistencies in the reported significance and in the treatment of lunar laser ranging constraints. The significance for the field is therefore conditional: the manuscript is a useful parameter study, but its headline conclusion needs additional robustness work before it can be regarded as a secure detection.","major_comments":[{"comment":"The headline BDG detection is not robust to the choice of low-z BAO data. With P18+DESI the Galileon parameter is 1/eα8 = 0.255+0.095−0.064 and H0 = 71.0+1.5−1.3, but with P18+(DESI+SDSS), which replaces the BGS and first LRG bins with SDSS low-z data, the same parameter becomes 0.157+0.088−0.076 (consistent with zero at about 2σ) and H0 = 69.2+0.9−1.2, with the SH0ES tension growing to 2.6σ. Since the abstract itself states that the results are 'mainly driven by the first two redshift bins of DESI,' the central claim is contingent on those two bins being free of systematics. The authors should either provide an explicit systematic-robustness analysis of the BGS and LRG1 measurements, or substantially rephrase the conclusions to present the BDG preference as dataset-dependent rather than as a detection.","section":"Abstract; Section IV.B; Table III"},{"comment":"The claim that 1/eα8 = 0.255+0.095−0.064 is 'about 3σ away from the ΛCDM value of 0' is not supported by the quoted asymmetric 68% interval alone: the ratio of the mean to the two one-sided errors gives values between about 2.7 and 4.0, and for a strongly non-Gaussian posterior either number is not a valid significance. The authors should report the posterior probability at zero, or the credible interval excluding zero, to justify the significance statement. This is load-bearing because the '3σ detection' is the paper's central quantitative result.","section":"Section IV.B, Eq. following 'about 3σ'"},{"comment":"The quoted time derivative of the cosmological gravitational constant for P18+DESI, Gdot/Gcosm(z=0) = (−11.1+2.4−3.9)×10^-13 yr^-1, does not contain zero at even the 2σ level under a Gaussian interpretation, and the same is true for the P18+(DESI+SDSS) value (−7.1+3.1−3.9)×10^-13. The sentence that these values are 'consistent with 0 at the 2σ level' is therefore inaccurate. The subsequent arguments that LLR constraints may be weakened by core-rotation correlations or by inhomogeneous local evolution are plausible directions but are not quantified here. As written, the paper simultaneously acknowledges exceeding the LLR limits and asserts consistency with zero, which is internally inconsistent and leaves the viability of BDG unclear.","section":"Section IV.B, Gdot/Gcosm paragraph; Table III"},{"comment":"In the BDG analysis the nonminimal coupling is fixed to ξ = 5×10^-5 rather than sampled, and only the Galileon amplitude 1/eα8 varies. Since the posterior for 1/eα8 is strongly degenerate with H0 and the model's viability depends on the chosen ξ, the reported 'detection' is conditional on an untested prior choice. A robustness check varying ξ within a range consistent with solar-system and cosmological constraints, or a discussion of why the result is insensitive to ξ, is needed before the 3σ claim can be taken as a property of the model class rather than of the chosen parameter point.","section":"Section III.B (BDG sampling)"},{"comment":"There is a numerical inconsistency in the EMG-CC results: the text states that with P18+DESI+SH0ES the 95% upper bound is 'V0 < 1.2,' while Table IV reports 'V0 < −1.2 (95%)'. The same table also uses an em-dash for the other dataset combinations, so the reader cannot tell which value is correct. In addition, the Conclusions section contains an incomplete sentence 'Δχ2 = − for EMG-CC' with a missing number. These need to be corrected.","section":"Table IV vs Section IV.D.3"}],"minor_comments":[{"comment":"The text says 'the 65% CI for the Hubble constant' where the context and tables indicate a 68% credible interval.","section":"Section IV.C"},{"comment":"The paper would be easier to reproduce if the chains or the modified CLASSig likelihood code were made public, or if a link to the existing public code were provided. The current reference list names Cobaya and CLASS, but not the exact version or repository used for the modified gravity extension.","section":"Section III (datasets)"},{"comment":"The definition of wDE in Eq. (C12) is clear, but the comparison with the w0waCDM curve in Figure 2 would benefit from a statement about how the latter is normalized (e.g., whether it is evaluated at the same best-fit background) and from error bars on the wDE curve, since the differences between models are otherwise hard to assess.","section":"Figure 2 and Appendix C"},{"comment":"The statement that DESI constrains 1/eα8 'by a factor of 2 better' than the earlier SDSS upper limit is imprecise; the earlier work gave an upper limit, not a measurement, so a direct comparison of precision is not straightforward. Please clarify what is meant.","section":"Section IV.B, previous constraint"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within the scope of a cosmology journal and the MCMC analysis appears technically sound. The main concerns are the dataset-dependence of the central claim, the questionable '3σ' and 'consistent with 0 at 2σ' statements, and the internal inconsistency in the EMG-CC V0 bound. These are fixable with additional analysis and revised wording, so I recommend major revision rather than rejection. I would encourage the editor to ask the authors to provide the posterior-based significance for the BDG detection and to include a quantitative discussion of the LLR tension."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe thing to know: this is a solid constraints paper whose headline claim—a ~3σ preference for the Galileon term in BDG and a reduction of the Hubble tension to 1.2σ—is explicitly fragile. The authors themselves show that swapping the two lowest DESI bins for SDSS data drops the Galileon signal to ~2σ and the Hubble tension to 2.6σ. So the interesting result is real but contingent, and the paper says so.\n\nWhat's new: applying the DESI 2024 BAO data to four scalar-tensor models, the BDG parameter 1/eα8 moves from an upper limit to a nonzero measurement. That is a genuine new inference, and the paper does it with a full Boltzmann treatment in the Jordan frame rather than a simplified background approximation. The cross-checks with DESI+SDSS, SH0ES, and CCHP priors make the dataset-dependence easy to see. The reporting is refreshingly direct—no overclaiming of what the data can do.\n\nSoft spots: the central claim rests on the first two DESI bins (BGS and LRG1). If those carry systematics, the headline collapses. The paper does not resolve this, only discloses it. That is the main caveat, not a hidden flaw. Minor issues: the '3σ' phrasing is a bit generous for an asymmetric credible interval that is consistent with zero at ~2σ under a different dataset choice; the fixed ξ = 5×10−5 is a parameter choice that deserves scrutiny; the likelihood code is not public, so the internal implementation cannot be independently verified. The BDG model also predicts a present-day Ġ/G that exceeds current LLR limits—the authors acknowledge this and invoke screening, which is plausible but not demonstrated here.\n\nOverall: the math and statistics appear internally consistent; the fragility is empirical rather than computational. This is exactly the kind of paper a serious referee should engage with, not desk-reject. I would not let the 1.2σ tension claim propagate without the code and a careful look at the low-z bins, but the paper earns a fair review.","headline":"A competent, honest constraints paper whose BDG 'detection' is explicitly contingent on the two lowest DESI redshift bins; worth refereeing, not worth treating as a discovery yet.","tokens_in":28897,"tokens_out":1646,"would_cite":true,"duration_ms":16107,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["98.80.-k","95.36.+x","04.50.Kd"],"model":"deepseek-v4-flash","headline":"The Brans-Dicke Galileon model, fitted to Planck and DESI 2024 BAO data, puts $H_0$ at 71.0 and brings the SH0ES tension down to $1.2\\sigma$.","keywords":["scalar-tensor gravity","Brans-Dicke Galileon","Hubble tension","DESI BAO","baryon acoustic oscillations","modified gravity","dark energy equation of state","Planck CMB"],"falsifier":"Recompute the P18+DESI analysis using pre-reconstruction BAO distances or the linear-point estimator for the BGS and LRG1 bins; if the Galileon parameter $1/e_{\\alpha 8}$ then becomes consistent with zero at below $2\\sigma$, the claimed $3\\sigma$ detection and the $1.2\\sigma$ Hubble-tension reduction would be shown to depend on reconstruction systematics in those two bins.","tokens_in":27727,"feed_emoji":"🌌","tokens_out":10545,"duration_ms":86020,"temperature":0.7,"pith_summary":"The paper argues that the 2024 DESI baryon acoustic oscillation measurements, combined with Planck CMB data, favor scalar-tensor modifications of gravity over a cosmological constant. In the Brans-Dicke Galileon model the preferred value of the Galileon parameter is $1/e_{\\alpha 8} = 0.255^{+0.095}_{-0.064}$, about $3\\sigma$ away from the $\\Lambda$CDM limit of zero, and the inferred Hubble constant rises to $H_0 = 71.0^{+1.5}_{-1.3}$ km/s/Mpc. At that value the long-standing $4.5\\sigma$ disagreement between early- and late-universe $H_0$ measurements drops to $1.2\\sigma$ with respect to the SH0ES local measurement. A sympathetic reader should care because, if the preference holds, a simple one-parameter extension of gravity can resolve the most prominent observational tension in cosmology while also producing the phantom dark-energy behavior that DESI has reported.","feed_headline":"Scalar-tensor model pushes H0 to 71 and eases Hubble tension to 1.2σ","feed_subtitle":"In Brans-Dicke Galileon gravity, Planck plus DESI BAO data favor a nonzero Galileon term at about 3σ.","key_machinery":"The central object is the Brans-Dicke Galileon (BDG) model, a scalar-tensor theory with a nonminimal coupling $F(\\sigma)R$, a negative kinetic term, and a Galileon term controlled by the parameter $\\alpha$; the sampled combination is written as $1/e_{\\alpha 8} = 10^{-8} e_\\alpha$, with the $\\Lambda$CDM limit at zero. In this theory the Galileon parameter is degenerate with the Hubble constant, so larger $1/e_{\\alpha 8}$ pushes $H_0$ upward while leaving most other cosmological parameters close to their $\\Lambda$CDM values. The model also relies on Vainshtein screening, which restores general relativity and the Newtonian gravitational constant on small scales while allowing the cosmological gravitational constant to evolve in time. The comparison models (induced gravity, $\\Delta$IG, and EMG-CC) isolate which feature, nonminimal coupling versus a Galileon term, is responsible for the data preference.","core_discovery":"On the paper's own terms, the discovery is that DESI 2024 BAO data break a previous degeneracy: the combination P18+DESI moves the Brans-Dicke Galileon parameter away from its $\\Lambda$CDM value of zero, giving $1/e_{\\alpha 8} = 0.255^{+0.095}_{-0.064}$ (68% CI, about $3\\sigma$), and correspondingly raises $H_0$ to $71.0^{+1.5}_{-1.3}$ km/s/Mpc. The paper shows this is driven mainly by the first two DESI redshift bins (BGS and LRG1) and that the same mechanism operates in induced gravity, induced gravity with a gravitational-constant offset, and early modified gravity, all of which prefer larger modified-gravity parameters and higher $H_0$ with DESI than with previous SDSS BAO data. In BDG the fit improves over $\\Lambda$CDM by $\\Delta\\chi^2 = -5.6$ ($\\Delta\\mathrm{AIC} = -3.6$), the dark-energy equation of state becomes phantom at low redshift, and the result sits between the SH0ES and CCHP local measurements, consistent with both.","pith_inferences":["If the $3\\sigma$ detection survives scrutiny of the low-redshift DESI bins, a next testable signature is the predicted present-day time variation of the gravitational constant, $\\dot{G}_{\\rm cosm}/G_{\\rm cosm} \\simeq -1.1 \\times 10^{-12}$ yr$^{-1}$, which improved lunar laser ranging or solar-system ephemerides could detect or rule out.","The paper's own dataset substitution suggests a decisive cross-check: reanalyzing with pre-reconstruction BAO estimators or the linear-point method in the BGS and LRG1 bins should either confirm or erase the $1/e_{\\alpha 8}$ detection.","Because only BDG among the four models generates a phantom equation of state, the comparison implies that the Galileon term, not the nonminimal coupling per se, is what makes the fit to DESI BAO data work; models without that term only mildly ease the tension."],"forward_implications":["If the BDG preference is real, Planck and SH0ES measurements of $H_0$ agree at the $1.2\\sigma$ level, so the Hubble tension does not require unknown systematics in either experiment.","The nonzero $1/e_{\\alpha 8}$ value implies a dark-energy equation of state that crosses below $-1$ at low redshift, aligning with DESI's preference for dynamical dark energy over a pure cosmological constant.","Constraints on the nonminimal coupling $\\xi$ in induced gravity become weaker with DESI than with SDSS because DESI pulls $H_0$ higher, changing the upper limits cosmology can place on modified-gravity parameters.","Replacing the two lowest DESI redshift bins with SDSS low-$z$ data reduces the Galileon significance from $3\\sigma$ to $2\\sigma$ and raises the $H_0$ tension to $2.6\\sigma$, showing that the result depends on those specific bins.","When a SH0ES prior is added, BDG improves over $\\Lambda$CDM by $\\Delta\\chi^2 = -23.4$ and $\\Delta\\mathrm{AIC} = -21.4$, making the model statistically preferred when local $H_0$ information is included."],"supporting_citations":[{"why":"DESI 2024 BAO measurements from galaxies, quasars, and Lyman-$\\alpha$ forest; the new dataset whose low-redshift bins drive the preference.","marker":"[5-7]"},{"why":"Planck 2018 CMB temperature, polarization, and lensing data used as the early-universe baseline (P18).","marker":"[52]"},{"why":"SH0ES local measurement $H_0 = 73.04 \\pm 1.04$ km/s/Mpc, the reference against which the Hubble tension is evaluated.","marker":"[53]"},{"why":"Previous analysis of the same BDG model with SDSS BAO, which yielded only an upper limit and serves as the direct baseline for the new detection.","marker":"[87]"},{"why":"DESI 2024 VI cosmological constraints, including the $w_0w_a$CDM phantom dark-energy fit and the composite DESI+SDSS dataset used in the robustness test.","marker":"[7]"},{"why":"CCHP local $H_0$ measurement that the BDG result must be consistent with.","marker":"[130]"},{"why":"Lunar laser ranging bound on $\\dot{G}/G$ used to test the model's predicted time variation of the gravitational constant.","marker":"[140]"},{"why":"Discussion of coupled Galileon challenges from local-Universe observations and screening suppression of $\\dot{G}$.","marker":"[100]"}],"fun_headline_variants":["DESI BAO data boost nonzero Galileon term, raising H0 to 71","Brans-Dicke Galileon gains 3σ support from DESI BAO","DESI BAO tilt gravity models toward higher H0","DESI data ease Hubble tension to 1.2σ in Galileon gravity"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The central claim stands on the assumption that the DESI BAO measurements in the first two redshift bins (BGS and LRG1, $z<0.8$) are free of systematics that mimic a preference for higher $H_0$; replacing those bins with SDSS low-$z$ data lowers the Galileon parameter from $3\\sigma$ to $2\\sigma$ significance and raises the Hubble tension from $1.2\\sigma$ to $2.6\\sigma$.","fun_headline_variants_meta":{"raw":{"variants":["DESI BAO data boost nonzero Galileon term, raising H0 to 71","Brans-Dicke Galileon gains 3σ support from DESI BAO","DESI BAO tilt gravity models toward higher H0","DESI data ease Hubble tension to 1.2σ in Galileon gravity"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000604,"raw_usage":{"total_tokens":2885,"prompt_tokens":1081,"completion_tokens":1804,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":697,"completion_tokens_details":{"reasoning_tokens":1721}},"tokens_in":697,"tokens_out":1804,"duration_ms":11311,"temperature":1.0,"reasoning_tokens":1721,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T14:24:42.298764+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute the P18+DESI analysis using pre-reconstruction BAO distances or the linear-point estimator for the BGS and LRG1 bins; if the Galileon parameter $1/e_{\\alpha 8}$ then becomes consistent with zero at below $2\\sigma$, the claimed $3\\sigma$ detection and the $1.2\\sigma$ Hubble-tension reduction would be shown to depend on reconstruction systematics in those two bins.","supporting_citations":[],"review_version":1}