{"id":"8e1e8b8e-6b9f-4fa1-bad2-5b0e5ce1c63d","arxiv_id":"2501.11619","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"A conformal dark U(1)' sector undergoing a strongly supercooled first-order phase transition can fit the NANOGrav 15-year gravitational wave background.","lead":"A model with a new dark force and a supercooled phase transition can produce the ripples in spacetime that pulsar timing arrays recently detected. The result overturns a recent argument that such phase transitions cannot explain the NANOGrav signal.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Omitted portal-induced mass term in the σ effective potential may shift the NANOGrav fit.","rationale":"The reader's weakest assumption is the full 1D (σ-only) treatment, including finite-temperature corrections and mixing with the Higgs. The present concern is narrower and more directly testable: even if two-field dynamics are negligible, the tree-level portal term with the Higgs vev produces a static mass correction to σ that appears to be omitted from the effective potential. The magnitude is small (about 2% of Mh2² at the upper bound of |λσh|), but because the central claim is a fit and the bounce action is exponentially sensitive, the omission is not automatically negligible. This is an internal consistency check: the paper asserts negligibility but does not demonstrate it. It does not require new physics or code beyond a rerun of CosmoTransitions with one extra term. The paper does independently verify percolation and uses standard templates, so the concern is not fatal until tested. The reader's CONDITIONAL verdict remains appropriate; the condition should include the check described above. Hence the verdict is unchanged.","tokens_in":7656,"tokens_out":39339,"duration_ms":407227,"concrete_test":"Recompute the σ-direction bounce for the best-fit point (gL = 0.59, Mh2 = 12.4 MeV, MZ' = 107.3 MeV, Table I) with ΔV = λσh v² φ²/4 added to the effective potential, for λσh = ±10⁻¹⁰, keeping all other inputs fixed. Compare S3(T), Tp, α, and β/H with the published values; if any change by more than ~10%, or if the predicted h²Ω_GW(f) peak moves outside the NANOGrav 95% band, the omitted portal term is physically relevant and the fit as presented is not robust. A broader check is to rerun the parameter scan with the portal term included and overlay the resulting 68% CL contours on Fig. 5.","verdict_should_be":"UNCHANGED","load_bearing_attack":"After electroweak symmetry breaking, the tree-level coupling λσh in Eq. (1) generates a quadratic term for the dark scalar: ΔV = λσh (v²/4) φ², where φ is the radial σ mode and v = 246 GeV. The paper's 'RG-improved thermal potential' section states that the σ direction decouples and that the role of λσh is rendered insignificant by the quadratic Higgs term, but it does not include this ΔV in the σ-direction potential used for the bounce. For the stated bound |λσh| < 10⁻¹⁰, |ΔV| corresponds to m_portal² ≈ 3 MeV². At the best-fit point (Table I: Mh2 = 12.4 MeV), this is ~2% of the scalon mass squared. Because the nucleation rate depends on S3/T exponentially, a 2% change in the mass term can shift S3/T by an amount of order Δ(S3/T) ~ 0.02 × 140 ≈ 3 at nucleation, which changes the percolation temperature and the derived α and β/H by an amount that is not bounded in the paper. The bounce action, percolation temperature, and gravitational wave spectrum are therefore not robust against the maximal allowed λσh. Since the central claim is a fit to NANOGrav data, neglecting this term without a quantitative statement is a load-bearing gap. A separate but related two-field issue (the reader's concern) is partly covered by this; here the effect is a static, one-dimensional correction that can be tested directly.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper claims that supercooled first-order phase transitions in a classically conformal dark U(1)' sector can explain the nHz stochastic gravitational wave background reported by NANOGrav, contrary to recent claims that such transitions fail to complete or reheat to too high temperatures. The authors construct a model with a dark scalar σ, a dark gauge boson Z', and a tiny Higgs-portal coupling, and compute the phase transition dynamics using a one-loop Coleman-Weinberg potential, thermal corrections, Daisy resummation, and the CosmoTransitions bounce package. They then fit the resulting gravitational wave spectrum to the NANOGrav 15-year data using PTArcade, obtaining preferred values MZ' in the 10-100 MeV range, gL ~ 0.6, and |λσh| < 1e-10. The paper argues that conformal models evade the no-go concerns of Ref. [9] because percolation is always possible and the reheating temperature stays below the dark-sector mass scale.","tokens_in":8093,"tokens_out":4730,"duration_ms":51816,"significance":"If the calculation is correct, the paper provides a concrete counterexample to the lore that supercooled FOPTs cannot explain the NANOGrav signal, and it identifies a falsifiable parameter corner of a conformal dark sector. The analysis uses standard and largely accepted tools (one-loop CW potential, thermal integrals, Daisy resummation, CosmoTransitions, PTArcade) and explicitly checks percolation, potential boundedness, and the BBN bound on dark radiation. The paper does not ship code or scan data, which limits reproducibility; the confidence regions are therefore not independently checkable from the manuscript alone.","major_comments":[{"comment":"The tree-level portal coupling λσh in Eq. (1) generates a Higgs-induced mass term for the dark scalar after electroweak symmetry breaking, ΔV = λσh (v²/4) φ². This term is not included in the σ-direction effective potential used for the bounce. For the allowed range |λσh| < 1e-10, one obtains m_portal² ≈ 3 MeV², which at the best-fit point (Table I, Mh2 = 12.4 MeV) is about 2% of the scalon mass squared. Because the nucleation action at Tn is S3/T ≈ 140 (Fig. 1), a 2% change in the mass term can shift S3/T by a few units, which in turn changes the percolation temperature Tp and the derived α and β/H values. The paper's statement that the impact of λσh is 'negligible' is not supported by a quantitative estimate, so the robustness of the NANOGrav fit within the allowed parameter range is not established. Please include this term in the bounce calculation or give a bound on the resulting shift in the SGWB spectrum.","section":"RG-improved thermal potential, Eq. (1)"},{"comment":"The reduction to a one-field bounce along the σ direction is asserted rather than demonstrated. The sentence 'we only consider the σ direction to be relevant for the FOPT as it decouples from the Higgs direction due to the strong scale hierarchy Mh1 ≫ Mh2' does not by itself imply that the two-field action equals the single-field action to sufficient accuracy. A heavy field can still affect the bounce if the tunneling path curves in field space, and finite-temperature corrections involving the Higgs may lower the barrier. Given the exponential sensitivity of the nucleation rate to the action, a two-field correction of order unity could move the predicted spectrum outside the NANOGrav band. I request a two-field bounce calculation for at least the best-fit point and a small set of scan points, or an analytic estimate showing that the deviation is negligible.","section":"RG-improved thermal potential"},{"comment":"The scan imposes the conditions TRH < M and α ≲ 1e8 in order to keep the signal in the nHz band. These are selection cuts rather than derived properties of the model. The claim that the model 'easily explains' NANOGrav is therefore a statement about the post-cut parameter space; the fraction of the unconstrained parameter space that satisfies these cuts is not quantified, and the priors are not physically motivated beyond the desire to match NANOGrav. Please clarify whether these conditions are generic for the model or define a fine-tuned region, since this bears directly on the 'easily' in the central claim.","section":"Numerical results"}],"minor_comments":[{"comment":"There is a duplicated word: 'supported supported by the Center...' should read 'supported by the Center...'.","section":"Acknowledgments"},{"comment":"The phrase 'the σ†σ term is absent' is imprecise: what is absent is the σ mass term, since the quartic λσ(σ†σ)² is present in Eq. (1). Please rephrase.","section":"Introduction"},{"comment":"The notation h²ΩGW uses h for the reduced Hubble constant, but h² is not explicitly defined before this equation; a one-line definition would help.","section":"Equation (5)"},{"comment":"The text says 'NANOGrav data favor Z′ and h2 masses in the 10–100 MeV range at 68% CL', while the Summary says 'dark sector masses of O(1−100) MeV'; please reconcile these ranges to avoid an apparent inconsistency.","section":"Numerical results"},{"comment":"The paper does not provide the scan data or a link to the code; given that the confidence regions are a central output, making the data available would substantially aid reproducibility and verification.","section":"Appendix"}],"recommendation":"major_revision","confidential_remarks":"The paper is interesting and likely correct in broad outline, but the two load-bearing gaps (the omitted portal-induced mass term and the unvalidated one-field approximation) need to be closed before publication. The lack of scan data and code is a reproducibility concern for a journal that publishes likelihood analyses. The selection cuts TRH < M and α ≲ 1e8 also deserve a clearer statement of whether they are generic or fine-tuned, as this affects the strength of the 'easily explains' claim."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear [Colleague],\n\nThe one thing to know: this letter is a serious counterexample to the recent lore (Athron et al.) that supercooled FOPTs can't explain the NANOGrav signal. Gonçalves et al. construct a dark U(1)' model with classical conformal symmetry only in the dark sector, reverse the mass hierarchy (Mh1 >> Mh2), and show that a strongly supercooled transition—with percolation completing and reheating below the dark sector scale—fits the NANOGrav 15-year data. The main claim is likely right.\n\nGenuinely good: they check percolation explicitly, check boundedness of the potential, use an RG-improved one-loop potential, and fit with PTArcade including the SMBHB contribution. The best-fit point (Mh2 ≈ 12 MeV, MZ' ≈ 107 MeV, gL ≈ 0.59, tiny portal) is concrete and testable by future PTAs. The construction is novel relative to their earlier conformal work: only the dark sector is conformal, which is what makes the reversed hierarchy workable. The citation pattern is fine; the heavy reliance on their own Ref. [10] is justified because that's where the framework lives.\n\nThe soft spots are two, both about what's left out of the sigma-direction potential. First, the portal coupling λσh generates a tree-level mass term for the dark scalar once EW symmetry breaks: ΔV = λσh v²/4 φ². The paper says this is insignificant but never includes it in the potential used for the bounce. For the stated bound |λσh| < 1e-10, this is ~2% of Mh2² at the best fit, and larger for lighter scalons in the allowed region. Because nucleation goes like exp(-S3/T), that shift can move the percolation temperature and the derived α and β/H by a non-negligible amount. This is a fixable gap, but it should be in the paper. Second, the reduction to a one-dimensional sigma direction is asserted, not demonstrated. With a 125 GeV Higgs and a 1e-10 portal, decoupling is plausible, but a two-field check would close it.\n\nNo code or scan data is shipped; for a letter that's the norm, but it does mean the numerical results aren't independently checkable. The cuts (TRH < M, |λσh| < 1e-10) are made with the NANOGrav band in mind; that's empirical fitting and they're transparent about it, not circularity.\n\nBottom line: this deserves a serious referee. Send it to review, and ask the authors to include the portal-induced mass term (or show it's negligible across the fitted range) and to briefly justify the sigma-only bounce. Those are moderate requests, not deal-breakers.","headline":"A concrete and likely-correct counterexample to the 'supercooled FOPTs can't explain NANOGrav' lore, with a fixable gap in the sigma-direction potential.","tokens_in":8627,"tokens_out":6027,"would_cite":true,"duration_ms":60620,"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":"A supercooled first-order phase transition in a conformal dark sector with a gauged U(1)' symmetry can explain the nanohertz gravitational-wave background observed by pulsar timing arrays.","keywords":["supercooled first-order phase transitions","conformal dark sector","gravitational wave background","NANOGrav","pulsar timing arrays","dark U(1)' gauge symmetry","Coleman-Weinberg mechanism","MeV-scale dark sector"],"falsifier":"Recompute the bounce action with the full two-field potential, including the Higgs direction, at the best-fit benchmark $g_L=0.59$, $M_{h_2}=12.4$ MeV, $M_{Z'}=107.3$ MeV. If including the second field changes the percolation temperature by more than an order of magnitude or prevents completion, the NANOGrav fit would not survive. Observationally, future pulsar timing data that show a pure power-law background continuing well below the predicted peak, with no turnover, would disfavor this explanation.","tokens_in":7499,"feed_emoji":"📡","tokens_out":12698,"duration_ms":118446,"temperature":0.7,"pith_summary":"This paper demonstrates that a strongly supercooled first-order phase transition in a conformal dark sector with a gauged U(1)' symmetry can produce the nanohertz gravitational-wave background seen by pulsar timing arrays. The demonstration matters because a previous no-go argument suggested such transitions either fail to complete or reheat the Universe too hot to leave a signal at those frequencies. In the model presented here, the transition completes, the reheating temperature stays below the dark-sector mass scale, and the predicted spectrum matches the NANOGrav 15-year data. If the claim holds, supercooled phase transitions remain viable explanations for pulsar timing array signals and come with a concrete, narrow dark-sector parameter region to target.","feed_headline":"Supercooled dark-sector transition fits the NANOGrav signal","feed_subtitle":"A supercooled conformal dark U(1)' transition reproduces the 15-year pulsar-timing signal","key_machinery":"The argument is carried by the RG-improved thermal effective potential along the dark scalar direction, $V_{\\rm eff}(\\phi_\\sigma, T)$, assembled from the tree-level potential, the one-loop radiative correction, thermal integrals, and Daisy resummation, with couplings and field renormalized at $\\mu=\\max[M_{Z'}(\\phi_\\sigma), \\pi T]$. The load-bearing object is the normalized Euclidean action $S_3/T$: in a conformal model it decreases monotonically toward zero as $T\\to 0$, allowing percolation to complete even for extreme supercooling, whereas in the nonconformal models behind the no-go argument the action is U-shaped and bounds the percolation temperature from below. Tunneling rates from this action yield the percolation temperature, reheating temperature, strength $\\alpha$, and inverse duration $\\beta/H$ that feed the gravitational-wave templates. The scalon, the light scalar that emerges from radiative symmetry breaking, together with the condition $T_{\\rm RH}<M_{Z'}$, keeps the peak frequency in the nanohertz band.","core_discovery":"The central claim is that a dark extension of the Standard Model in which only the dark sector is classically conformal, with radiative U(1)' breaking, produces a first-order phase transition that is strongly supercooled yet completes. Because the conformal potential barrier persists to zero temperature, the normalized Euclidean action $S_3/T$ falls toward zero as $T\\to 0$, so percolation can finish at very low temperature, and because the barrier is conformal the reheating temperature does not jump up to the dark-sector mass scale. The resulting gravitational-wave spectrum, built from an RG-improved thermal potential and current spectral templates, fits the NANOGrav 15-year signal with $Z'$ and scalon masses in the 10--100 MeV range, a gauge coupling $g_L\\simeq 0.59$, and a negligibly small portal coupling $|\\lambda_{\\sigma h}|<10^{-10}$. This is offered as a concrete counterexample to the claim that supercooled first-order phase transitions cannot explain pulsar timing array data.","pith_inferences":["The paper computes the transition along a single field direction; if two-field effects involving the Higgs are included, the favored region could shift or shrink. A natural next step, not taken in the paper, would be a full two-dimensional bounce calculation.","The mechanism is not obviously specific to U(1)': any classically conformal dark sector with a radiatively broken gauge symmetry could behave similarly, though the quantitative fit would differ. The paper's conclusion therefore suggests a broader family of PTA-compatible supercooled transitions.","Because the fit demands $|\\lambda_{\\sigma h}|<10^{-10}$, the dark scalar is essentially decoupled from the Higgs, so the gravitational-wave signal may be the only observable consequence of this sector; collider signatures would be extremely weak."],"forward_implications":["Supercooled first-order phase transitions are not generically ruled out as pulsar timing array explanations; the no-go argument applies to nonconformal models, not to conformal dark sectors.","At the best-fit point the supermassive black hole binary contribution is negligible, so the NANOGrav signal can be fully explained by the dark sector transition alone.","The favored parameter region---$Z'$ and scalon masses of order 10--100 MeV, $g_L$ near 0.6, $\\lambda_\\sigma$ of order $-0.01$, and $|\\lambda_{\\sigma h}|<10^{-10}$---provides concrete targets for nHz gravitational-wave experiments and dark-sector searches.","Making only the dark sector conformal, with $\\mu_h^2\\neq 0$, avoids the Higgs vacuum stability obstruction that blocks the alternative hierarchy."],"supporting_citations":[{"why":"This supplies the NANOGrav 15-year data that the model is fit to.","marker":"[1]"},{"why":"This shows that new-physics scenarios are statistically more compatible with the signal than supermassive black hole binaries alone, motivating the FOPT explanation.","marker":"[5]"},{"why":"This is the no-go argument that supercooled FOPTs may not complete and may reheat too hot; the paper's central task is to circumvent it.","marker":"[9]"},{"why":"This defines the conformal dark U(1)' framework and the RG-improved thermal potential used here.","marker":"[10]"},{"why":"This is the Coleman-Weinberg radiative symmetry breaking mechanism that generates the potential barrier and the scalon.","marker":"[14]"},{"why":"This supports the claim that reheating in a conformal U(1) model stays well below the gauge boson mass scale.","marker":"[22]"},{"why":"This provides the bounce-action code used to compute tunneling rates and thermodynamic parameters.","marker":"[25]"},{"why":"This supplies the spectral templates for bubble collisions and sound waves used to convert thermodynamic parameters into gravitational wave spectra.","marker":"[26]"},{"why":"This supplies the statistical fitting package used to derive confidence regions from the NANOGrav 15-year dataset.","marker":"[30]"},{"why":"This provides the likelihood configuration used inside the statistical fitting package.","marker":"[31]"}],"fun_headline_variants":["Dark supercooled transition resolves NANOGrav puzzle","Conformal dark sector explains NANOGrav's 15-year signal","Supercooled conformal dark transition matches NANOGrav","NANOGrav signal from a conformal dark transition","Supercooled dark U(1)' phase transition fits NANOGrav"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The calculation assumes the phase transition runs purely along the dark scalar field, with the ordinary Higgs field completely decoupled because the Higgs is much heavier than the dark scalar and its interaction with it is extremely weak.","fun_headline_variants_meta":{"raw":{"variants":["Dark supercooled transition resolves NANOGrav puzzle","Conformal dark sector explains NANOGrav's 15-year signal","Supercooled conformal dark transition matches NANOGrav","NANOGrav signal from a conformal dark transition","Supercooled dark U(1)' phase transition fits NANOGrav"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000889,"raw_usage":{"total_tokens":3764,"prompt_tokens":805,"completion_tokens":2959,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":421,"completion_tokens_details":{"reasoning_tokens":2870}},"tokens_in":421,"tokens_out":2959,"duration_ms":25685,"temperature":1.0,"reasoning_tokens":2870,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T18:03:44.708365+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute the bounce action with the full two-field potential, including the Higgs direction, at the best-fit benchmark $g_L=0.59$, $M_{h_2}=12.4$ MeV, $M_{Z'}=107.3$ MeV. If including the second field changes the percolation temperature by more than an order of magnitude or prevents completion, the NANOGrav fit would not survive. Observationally, future pulsar timing data that show a pure power-law background continuing well below the predicted peak, with no turnover, would disfavor this explanation.","supporting_citations":[],"review_version":1}