{"id":"d2edf872-5b2b-427d-9b05-831155230303","arxiv_id":"2507.18188","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Analysis of EPTA DR2 timing data yields Gdot/G = (0.32 ± 0.31)e-12 yr^-1 and kappa_D = (-0.04 ± 0.14)e-4 at 95% confidence for PSR J1713+0747, consistent with general relativity.","lead":"This paper uses updated European Pulsar Timing Array data to measure how fast the orbits of three binary pulsars are shrinking, then translates those measurements into limits on two possible violations of general relativity: a slowly changing gravitational constant and the emission of dipole gravitational waves. The tightest new bound, from PSR J1713+0747, is consistent with Einstein's theory and improves the previous limit on dipole radiation by about an order of magnitude.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Face-value use of Eq. (3.26) cannot yield the quoted κD bound: propagating the stated 5.7×10^-19 s^-1 uncertainty of PSR J0437–4715 through Eq. (3.23) gives δκD ~ 0.2, roughly 10^4 times the reported 1.4×10^-5.","rationale":"The reader's weakest assumption, the neutron-star sensitivity relation Eq. (2.15), is a valid structural input and does affect κD through sp^2. However, the single most load-bearing concern is more direct: the paper's own quoted auxiliary pulsar input, Eq. (3.26), is numerically incompatible with the reported κD constraint, Eq. (4.2). The order-of-magnitude mismatch is 10^3-10^4 and cannot be absorbed by plausible variations in masses or sensitivities, so it points either to a mislabeled quantity in Eq. (3.26) or to a failure to propagate the auxiliary pulsar's uncertainty. The test I propose will settle which. If the first alternative is correct, a corrected version of the paper could preserve the qualitative conclusion; if the second is correct, the headline κD bound is invalid and must be recomputed. For this reason I keep the verdict CONDITIONAL rather than moving to REJECT, but the condition is explicit: the J0437 input must be correctly stated and propagated, and Eq. (4.2) must be checked against the result. The G_dot/G constraint and the pipeline description remain useful, and the paper is transparent about many limitations, but the κD centerpiece is not reproducible from the text as written.","tokens_in":17848,"tokens_out":31807,"duration_ms":339767,"concrete_test":"Decisive check: rerun the two-pulsar fit exactly as described in §3.3, treating the quantity in Eq. (3.26) as a Gaussian constraint with mean 3.2×10^-19 and standard deviation 5.7×10^-19 s^-1 on (Pb_dot/Pb)^exc for PSR J0437-4715, using the same masses, orbits, and Eq. (2.15), and report the resulting 95% interval for κD. Then repeat the same fit after reinterpreting Eq. (3.26) as an absolute Pb_dot, dividing by Pb = 4.96×10^5 s to obtain the fractional excess. If the first version gives δκD ≈ 0.2 and the second gives a result compatible with Eq. (4.2), Eq. (3.26) is mislabeled and the manuscript must be corrected. If neither version reproduces Eq. (4.2), the uncertainty was not propagated and the combined analysis must be redone before the κD headline can be used.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The headline κD constraint is not reproducible from the stated auxiliary pulsar input. In Eq. (3.23), the fractional excess is written as Pb_dot_exc/Pb = -2(G_dot/G)[1-(1+3Mc/2M)sp] - (4π^2 T⊙ Mc/Pb^2)[q/(q+1)]κD sp^2. For PSR J0437-4715, Pb = 5.74 d = 4.96×10^5 s, Mc ≈ 0.17 M⊙, q ≈ 8, and Eq. (2.15) gives sp ≈ 0.16, so the coefficient of κD in the last term is about 3×10^-18 s^-1. The paper's adopted input, Eq. (3.26), is (Pb_dot/Pb)^exc = (3.2 ± 5.7)×10^-19 s^-1. Propagating the stated 5.7×10^-19 uncertainty alone gives δκD ≈ 0.2, whereas Eq. (4.2) quotes κD = (-0.04 ± 0.14)×10^-4. Moreover, the face-value central value 3.2×10^-19 s^-1 would pull κD to about -0.1 when combined with the reported G_dot/G, not -4×10^-6, if it is included as a fractional excess in the two-equation solve described in §3.3. Therefore either Eq. (3.26) is mislabeled (for example, the quantity is an absolute Pb_dot rather than a fractional excess), or the uncertainty of the auxiliary pulsar was not propagated into Eq. (4.2). In either case, the reported κD bound and the central claim of the tightest pulsar limit on dipole radiation are unsupported as stated.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript presents a Bayesian re-analysis of pulsar timing data for PSR J1713+0747, PSR J1738+0333, and PSR J1012+5307 using the TEMPO2/MCMC4Tempo2 pipeline with recent EPTA DR2 data, and combines the inferred orbital-period derivatives with a literature value for PSR J0437–4715 to constrain a possible time variation of Newton's constant Gdot/G and the dipole radiation parameter κD. The central results are Gdot/G = (0.32 ± 0.31) × 10^-12 yr^-1 and κD = (-0.04 ± 0.14) × 10^-4 for the J1713+0747 combination, which the authors argue are consistent with GR and constitute the tightest pulsar limit on dipole radiation. The paper also provides a public pipeline and discusses theoretical interpretations in scalar-tensor and extra-dimensional frameworks.","tokens_in":18229,"tokens_out":12770,"duration_ms":122557,"significance":"If the analysis were correct, the κD constraint would be the most stringent pulsar-based limit on dipole gravitational radiation, improving on previous work by an order of magnitude, and would complement Solar System bounds on Gdot/G. The use of an end-to-end MCMC pipeline on public EPTA DR2 data is a reproducible methodology. However, the paper's headline result is not reproducible from the stated inputs, and the inconsistency is so large that the significance claim cannot be credited without a full reanalysis.","major_comments":[{"comment":"Equation (4.2) is not derivable from the stated inputs. For PSR J0437–4715, using Pb = 5.74 d, Mc ≈ 0.17 M⊙, q ≈ 8, and sp ≈ 0.16, the coefficient of κD in Eq. (3.23) is approximately 3 × 10^−18 s^−1. Propagating the uncertainty quoted in Eq. (3.26), δ(Pdot/Pb) = 5.7 × 10^−19 s^−1, gives δκD ≈ 0.2, whereas Eq. (4.2) reports an uncertainty of 1.4 × 10^−5. Even the central value in Eq. (3.26) would shift κD by about −0.1 if combined with the reported Gdot/G. Thus either Eq. (3.26) is mislabeled (e.g., an absolute Pdot rather than a fractional excess), or the uncertainty of the auxiliary pulsar was not propagated into the joint fit. In either case, the headline constraint on κD is unsupported and must be rederived or retracted.","section":"§3.3/§4, Eq. (3.23), Eq. (3.26), Eq. (4.2)"},{"comment":"The κD uncertainties quoted for PSR J1738+0333 and PSR J1012+5307 (of order 10^−7) are even tighter than the J1713+0747 result and require an implausibly small uncertainty in the fractional orbital-period derivative. For these systems the coefficient of κD in Eq. (3.23) is of order 10^−15–10^−16 s^−1 (for Pb ~ 0.35 d and 0.60 d), so the reported σ_κD ≈ 1 × 10^−7 would demand σ(Pdot/Pb) ≈ 1 × 10^−23 s^−1. The paper does not report the measured Pdot/Pb values or their uncertainties for any of the three pulsars, so this required precision is not demonstrated. A table of the fitted parameters and the derived fractional orbital-period derivatives with uncertainties is necessary to assess the claims.","section":"§3.3/§4, Eqs. (4.3)–(4.4)"},{"comment":"The paper quotes uncertainties as '95% confidence level' throughout Section 4, but Section 3.3 states that posterior means and standard deviations are computed from the MCMC chains. If the quoted numbers are 1σ values, the 95% intervals would be roughly twice as wide; if they are already 95%, they should not be labeled as standard deviations. This ambiguity directly affects the claimed agreement with General Relativity and with previous constraints, and it should be clarified with a consistent error-bar convention.","section":"§3.3/§4"}],"minor_comments":[{"comment":"The auxiliary pulsar is referred to as PSR J0437–4714 once; this should read PSR J0437–4715.","section":"§4, first paragraph"},{"comment":"The derivation of the reported constraints would be greatly facilitated by a table listing the fitted Keplerian and post-Keplerian parameters, especially Pdot/Pb, for each of the three main pulsars. Without these values, the two-equation solution of Eq. (3.23) cannot be independently checked.","section":"§3.3/§4"},{"comment":"The paper notes that O(sp^3) terms neglected in Eq. (2.17) may be important for some theories, but it does not quantify how the adopted linear sensitivity relation of Eq. (2.15) affects the final κD bound. Since the dipole term scales as sp^2, a brief systematic propagation of this assumption would strengthen the result's robustness.","section":"§5, last paragraph"},{"comment":"In Eq. (2.3) the Newton constant and orbital period appear without an explicit c^3 factor; the authors should state that geometrized units with c = 1 are used throughout, or include the corresponding factors, to avoid dimensional confusion.","section":"§2, Eq. (2.3)"}],"recommendation":"major_revision","confidential_remarks":"The discrepancy identified in Major Comment 1 is severe: the reported κD uncertainty is four orders of magnitude smaller than the error propagation from the stated auxiliary pulsar input allows. If the authors cannot trace a typo in their propagation code or in Eq. (3.26), the main result will collapse. I would request the actual Pdot/Pb values for all pulsars and a step-by-step propagation before reconsidering the paper."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThis is a dataset update with a numerical inconsistency at its core. The EPTA DR2 timing data and the choice to leave kappa_D free for the auxiliary pulsar are both reasonable, but the headline constraint quoted in Eq. (4.2) does not follow from the stated inputs. Propagate the auxiliary pulsar's stated uncertainty through Eq. (3.23) and you get sigma(kappa_D) ~ 0.2, roughly 10^4 times larger than the reported 1.4e-5. Either Eq. (3.26) is mislabeled, or the uncertainty was never propagated. This is the main result, not a side detail.\n\nThe paper does have merit. The timing pipeline is established, the J1713+0747 fit itself appears sound, and the authors are candid about limitations: they note the O(sp^3) neglect, flag discrepancies in J1738 and J1012, and point out that the sensitivity relation is a linear-in-mass approximation. The code release promise is a plus, though the repo is not yet public.\n\nThe soft spots beyond the arithmetic: the J1738 and J1012 constraints on kappa_D come out 100 times tighter than the J1713 one, which the authors themselves treat as untrustworthy—that is another sign the error budget is not under control. Galactic acceleration uncertainties are not fully propagated. And the auxiliary pulsar's 2008 literature value carries a huge uncertainty relative to what the final result claims.\n\nBottom line: this deserves a serious referee, but as submitted it should be sent back for major revision. The data processing is worth reporting; the claimed kappa_D bound is not supported. As written, I would not cite it.","headline":"Updated dataset, standard pipeline, but the quoted kappa_D uncertainty is inconsistent with the paper's own equations by about four orders of magnitude.","tokens_in":18799,"tokens_out":15881,"would_cite":false,"duration_ms":148137,"reading_group":"maybe","serious_thinker":"no","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Timing of two pulsars gives the tightest pulsar bound yet on dipole gravitational radiation and finds the gravitational constant steady.","keywords":["binary pulsars","modified gravity","gravitational constant variation","dipolar gravitational radiation","pulsar timing","scalar-tensor theories","Markov chain Monte Carlo","general relativity tests"],"falsifier":"Replace the adopted sensitivity $s_p = 0.16(M_p/1.33 M_\\odot)$ with a value from a different equation of state or from a full scalar-tensor neutron-star model for the same masses and recompute the joint fit; if the inferred $\\kappa_D$ moves by much more than the quoted $0.14 \\times 10^{-4}$, the reported bound is set by the assumed neutron-star physics rather than by the timing data. A second check would be to add another low-eccentricity pulsar--white-dwarf binary with a well-measured parallax and an orbital period between roughly 2 and 20 days: the joint posterior for $\\dot{G}/G$ and $\\kappa_D$ should shift along the expected $P_b$ versus $P_b^{-2}$ direction if the degeneracy-breaking is working as claimed.","tokens_in":17621,"feed_emoji":"⏱️","tokens_out":10962,"duration_ms":108329,"temperature":0.7,"pith_summary":"Binary pulsars are among the best natural laboratories for modified gravity, because two proposed deviations — a slowly varying gravitational constant $G$ and dipole gravitational radiation — would show up as an extra decay of the orbital period. This paper fits recent high-precision timing data for three pulsars without assuming a theory of gravity, corrects the apparent orbital decay for Galactic acceleration and the geometric acceleration from transverse motion, and then combines systems with different orbital periods to separate the two effects. For its best system, PSR J1713+0747, it obtains $\\dot{G}/G = (0.32 \\pm 0.31) \\times 10^{-12}~\\mathrm{yr}^{-1}$ and $\\kappa_D = (-0.04 \\pm 0.14) \\times 10^{-4}$ at 95% confidence. The paper concludes that the results are consistent with general relativity, with the dipole bound an order of magnitude tighter than the earlier J1713+0747 limit. If correct, this narrows the parameter space for scalar-tensor and related modified-gravity theories while confirming that the gravitational constant is steady at the $10^{-12}$ per year level.","feed_headline":"Pulsar pair finds gravity constant steady in tightest dipole test","feed_subtitle":"At 95% confidence, Gdot/G = (0.32 ± 0.31) × 10^-12 yr^-1 and the dipole parameter is (-0.04 ± 0.14) × 10^-4.","key_machinery":"The load-bearing relation is Eq. (3.23): $\\dot{P}_b^{\\mathrm{exc}}/P_b = -2(\\dot{G}/G)[1 - (1 + 3M_c/2M)s_p] - 4\\pi^2 T_\\odot M_c/P_b^2 \\, q/(q+1) \\, \\kappa_D s_p^2$, where $M$ is the total mass, $q=M_p/M_c$, $T_\\odot$ is the solar-mass time unit, and $s_p$ is the neutron-star sensitivity, approximated as linear in mass, $s_p = 0.16 (M_p/1.33 M_\\odot)$. The two terms scale differently with orbital period — the $\\dot{G}$ term grows with $P_b$ while the dipole term falls as $P_b^{-2}$ — so fitting two binaries with different periods breaks the degeneracy between the parameters. The paper feeds this relation with full posterior samples from the timing fits rather than single best-fit values, which propagates the parameter correlations into the final constraints.","core_discovery":"The central claim is that binary-pulsar timing can jointly constrain a time-varying gravitational constant $\\dot{G}/G$ and the dipole radiation parameter $\\kappa_D$, and that with current data both are statistically compatible with zero. The authors use Bayesian Markov chain Monte Carlo fits to the pulse arrival times to extract the observed orbital-period derivative, then subtract kinematic Doppler corrections and the general-relativistic quadrupole prediction to isolate any excess decay. Joining PSR J1713+0747 with the auxiliary pulsar PSR J0437-4715 — whose well-measured excess orbital-period derivative is taken from the literature — gives the 95% constraints quoted above. A distinctive modeling choice is that, unlike earlier work, the auxiliary pulsar's dipole term is not fixed to zero; keeping both parameters free avoids absorbing a possible dipole signal into the $\\dot{G}/G$ measurement. The reported values are consistent with general relativity, and the paper finds that imposing the older zero-dipole assumption changes the constraints only slightly.","pith_inferences":["Because $\\kappa_D$ enters only through $s_p^2$, the reported limit is effectively a joint constraint on gravity and on neutron-star structure; a future equation of state that changes $s_p$ by tens of percent would shift the dipole bound by a comparable factor even with identical timing data.","The comparison with earlier work suggests that part of the order-of-magnitude improvement in $\\kappa_D$ comes from the recent larger timing data sets rather than from a fundamentally new method; applying the same pipeline to older data would isolate how much of the gain is data-driven.","A natural next step would be to apply the same two-parameter fit to several short-period pulsar--white-dwarf binaries with well-measured parallaxes, since those systems weight the dipole term more heavily and would provide an independent cross-check of the degeneracy-breaking reported here.","If the assumption that the companion's sensitivity is negligible were relaxed for a white-dwarf companion, the $\\dot{G}/G$ term in Eq. (3.23) would acquire an additional mass-dependent correction; checking that this correction is below the reported uncertainty would test the stability of the quoted bound."],"forward_implications":["If the central result is correct, modified-gravity theories predicting dipole radiation stronger than about $|\\kappa_D| \\sim 10^{-5}$ are disfavored by this pulsar, independent of the specific scalar coupling.","The $\\dot{G}/G$ bound of $(0.32 \\pm 0.31) \\times 10^{-12}\\,\\mathrm{yr}^{-1}$ implies the effective gravitational coupling has been steady to roughly a part in $10^{12}$ per year over the timing baseline.","Because the dipole contribution scales as $P_b^{-2}$ and the $\\dot{G}$ contribution scales as $P_b$, the combined analysis of binaries with different orbital periods is the mechanism that lets a single pair of systems constrain both parameters at once.","The authors' choice to leave the auxiliary pulsar's dipole term free means the quoted limits are conservative against a hidden dipole signal; fixing the term to zero, as earlier work did, gives slightly weaker but still GR-consistent constraints.","The reported limits translate directly into a bound on the time evolution of the scalar field in scalar-tensor theories, since $\\dot{G}/G$ maps to $\\dot{\\phi}/\\phi$ in those frameworks."],"supporting_citations":[{"why":"supplies the definitions of gravitational sensitivities and the leading-order dipolar orbital-decay expression used in Eq. (2.17).","marker":"[31]"},{"why":"provides the result behind the adopted linear neutron-star sensitivity scaling with mass in Eq. (2.15).","marker":"[32]"},{"why":"gives the earlier J1713+0747 constraints on $\\dot{G}/G$ and $\\kappa_D$ against which the new bounds are compared.","marker":"[34]"},{"why":"establishes the combined two-pulsar method that separates $\\dot{G}/G$ from dipole radiation using binaries with different orbital periods.","marker":"[51]"},{"why":"provides the recent high-precision multi-telescope timing data sets from which the orbital and post-Keplerian parameters are fitted.","marker":"[60]"},{"why":"supplies the external precise excess orbital-period derivative for the auxiliary pulsar PSR J0437-4715 used in the joint analysis.","marker":"[61]"},{"why":"offers the comparison system PSR J1738+0333 analysis and points to the treatment of $O(s_p^3)$ terms in the dipole formula.","marker":"[62]"}],"fun_headline_variants":["Pulsar timing constrains G-variation and dipole radiation","Binary pulsars tighten limits on modified gravity","No drift in G: pulsar data pass GR test","Updated pulsar test: gravity constant unchanged","Pulsar pair probes gravity, finds GR intact"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The translation of the measured excess orbital decay into a dipole-radiation bound rests on a single scaling relation for how strongly a neutron star's mass responds to a change in the effective gravitational constant, adopted as linear in mass with a coefficient from one equation of state; because the dipole term is proportional to the square of that sensitivity, any error in the scaling changes the $\\kappa_D$ bound quadratically, and the paper also drops $O(s_p^3)$ terms it notes may matter in some theories.","fun_headline_variants_meta":{"raw":{"variants":["Pulsar timing constrains G-variation and dipole radiation","Binary pulsars tighten limits on modified gravity","No drift in G: pulsar data pass GR test","Updated pulsar test: gravity constant unchanged","Pulsar pair probes gravity, finds GR intact"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000323,"raw_usage":{"total_tokens":1877,"prompt_tokens":1073,"completion_tokens":804,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":689,"completion_tokens_details":{"reasoning_tokens":728}},"tokens_in":689,"tokens_out":804,"duration_ms":9598,"temperature":1.0,"reasoning_tokens":728,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T14:38:23.823434+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Replace the adopted sensitivity $s_p = 0.16(M_p/1.33 M_\\odot)$ with a value from a different equation of state or from a full scalar-tensor neutron-star model for the same masses and recompute the joint fit; if the inferred $\\kappa_D$ moves by much more than the quoted $0.14 \\times 10^{-4}$, the reported bound is set by the assumed neutron-star physics rather than by the timing data. A second check would be to add another low-eccentricity pulsar--white-dwarf binary with a well-measured parallax and an orbital period between roughly 2 and 20 days: the joint posterior for $\\dot{G}/G$ and $\\kappa_D$ should shift along the expected $P_b$ versus $P_b^{-2}$ direction if the degeneracy-breaking is working as claimed.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"supplies the definitions of gravitational sensitivities and the leading-order dipolar orbital-decay expression used in Eq. (2.17)."},{"cited_title":"Damour and G","cited_arxiv_id":null,"evidence_quote":"provides the result behind the adopted linear neutron-star sensitivity scaling with mass in Eq. (2.15)."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"gives the earlier J1713+0747 constraints on $\\dot{G}/G$ and $\\kappa_D$ against which the new bounds are compared."},{"cited_title":"Lazaridis, N","cited_arxiv_id":null,"evidence_quote":"establishes the combined two-pulsar method that separates $\\dot{G}/G$ from dipole radiation using binaries with different orbital periods."},{"cited_title":"Antoniadis et al","cited_arxiv_id":null,"evidence_quote":"provides the recent high-precision multi-telescope timing data sets from which the orbital and post-Keplerian parameters are fitted."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"supplies the external precise excess orbital-period derivative for the auxiliary pulsar PSR J0437-4715 used in the joint analysis."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"offers the comparison system PSR J1738+0333 analysis and points to the treatment of $O(s_p^3)$ terms in the dipole formula."}],"review_version":1}