{"id":"9045782f-80a1-4f72-9ca9-c902637f0d71","arxiv_id":"2601.14412","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Slow first-order phase transitions generate comoving curvature perturbations too small to form primordial black holes, with a new fitting template for the power spectrum.","lead":"This paper re-evaluates curvature perturbations from slow, strongly supercooled first-order phase transitions using a gauge-invariant multi-fluid formalism. It finds the perturbations are too small to form primordial black holes and provides a fitting formula for future observational constraints.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The DeltaPT-to-linear-perturbation mapping is the weak link: the slicing ambiguity cancels at linear order, but δ_NG is not shown to be small, so nonlinear corrections could spoil the PBH suppression. A nonlinear separate-universe ζ check is needed.","rationale":"The reader correctly identifies the conversion from DeltaPT density contrasts to a gauge-invariant curvature perturbation as the loading-bearing step, and the paper itself flags the assumption in Sec. IV.A. My refinement is that the unknown time-slicing is less dangerous than it sounds: at linear order, the relative entropy perturbation S_Vr is invariant under a common time shift, so the cancellation in Eq. (18) protects the calculation from the unspecified slicing. The real danger is nonlinearity. The paper provides no quantitative evidence that δ_NG is small enough for linear perturbation theory, and because the final |R| is tiny, even O(δ²) corrections could be relevant. This validates the reader's CONDITIONAL verdict but shifts the emphasis from gauge ambiguity to the small-δ assumption. The proposed nonlinear separate-universe test would settle the issue directly and is a natural extension of the authors' own methodology.","tokens_in":12294,"tokens_out":10542,"duration_ms":121267,"concrete_test":"Implement a fully nonlinear separate-universe pipeline for the same DeltaPT realizations: for each patch, integrate the local Friedmann equation with ρ_V = F ΔV and the radiation continuity equation, then define the nonlinear curvature perturbation on uniform total-density slices by ζ_NL(t) = ln[a_loc(t_loc)/a_bg(t)], where t_loc is chosen so that ρ_loc(t_loc) = ρ_bg(t). Compare the variance σ²(ζ_NL) and the distribution tail at t_k with the linear-order result from Eqs. (13), (17), and (18). If σ²(ζ_NL) differs by more than ~50%, or if the inferred δ_C tail crosses the adopted δ_crit ≈ 0.45, the PBH suppression in Sec. VI is not robust.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central PBH-suppression claim rests on converting separate-universe density contrasts δ_NG into a gauge-invariant curvature perturbation via linear multi-fluid perturbation theory. The paper explicitly assumes in Sec. IV.A that δ_NG can be approximated, at linear order, as density contrasts on an unspecified time-slicing. The unknown slicing is not itself fatal: in the relative entropy perturbation S_Vr (Eq. 18), a linear time shift δt cancels in δρ_α/˙ρ_α − δρ_β/˙ρ_β, so the linear-order S_Vr is slicing-independent. What is unprotected is the linearization itself. The manuscript never shows δ_NG ≪ 1; Fig. 1 plots |δ_NG| and the reference |9/4 δ_NG| on an axis extending to order unity, and no convergence test in δ is reported. If δ_NG is only O(0.1–0.4), second-order corrections to the gauge transformation and to the mapping from a finite patch to a global perturbation enter S_Vr and δP_rel at O(δ²). The resulting |R| in Fig. 1 is orders of magnitude smaller (~10⁻⁴ or less), so even modest nonlinear corrections could change the sign or magnitude of the curvature perturbation and undo the claimed suppression relative to Ref. [24]. The conclusion that PBHs are unlikely is therefore not yet robust against the most obvious failure mode of the linearization.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper claims that super-horizon curvature perturbations generated by slow, strongly supercooled first-order phase transitions are much smaller than earlier separate-universe estimates suggested, so primordial black hole production from these perturbations is unlikely. The authors use the multi-fluid gauge-invariant perturbation formalism of Refs. [32,33], feeding in density contrasts δ_NG obtained from the DeltaPT separate-universe simulation of Ref. [24]. They compute the comoving curvature perturbation R via the relative entropy perturbation S_Vr, present its variance for several β/H_n and k/k_max, derive a fitting formula for P_R(k) (Eq. 20), and use this template to discuss constraints from CMB, Lyman-α, FIRAS, PTA, SKA, LISA, and UCMHs, as well as scalar-induced gravitational waves and a NANOGrav 15-year fit. The central conclusion is that the δ_C distribution lies far below the adopted PBH threshold δ_crit ≈ 0.45.","tokens_in":12689,"tokens_out":2721,"duration_ms":32497,"significance":"If the main claim is correct, the paper resolves a gauge ambiguity that has affected recent estimates of PBHs and SIGWs from slow FOPTs, and it provides a useful fitting formula for phenomenological studies. The use of the standard multi-fluid gauge-invariant formalism is appropriate, and the paper is explicit about its key assumption. The numerical analysis with 50,000 realizations per parameter point is a strength, as is the inclusion of current and projected observational constraints and a PTArcade-based posterior analysis. However, the central result is load-bearing on a linearization assumption that is not quantitatively validated, and the simulation data/code are not shipped, so the core suppression claim is not yet fully demonstrable to the standards of a journal publication.","major_comments":[{"comment":"The central result depends on the assumption, stated in Sec. IV.A, that the separate-universe density contrasts δ_NG can be approximated at linear order as density contrasts on a specific but unknown time-slicing. While a linear time shift cancels in S_Vr, the linearization itself is not justified: the manuscript nowhere shows δ_NG ≪ 1. Figure 1 plots |δ_NG| and |9/4 δ_NG| on an axis extending to order unity, and no convergence test in δ is reported. Since the final |R| values are orders of magnitude smaller (~10^-4 or less), second-order corrections to the gauge transformation and to the mapping from a finite patch to a global perturbation enter S_Vr and δP_rel at O(δ^2) and could change the sign or magnitude of R. The conclusion that PBHs are unlikely is therefore not robust against this most obvious failure mode. A quantitative bound on δ_NG in the relevant (β/H_n, k/k_max) regime, or","section":"Sec. IV.A; Eq. (18); Fig. 1"},{"comment":"No statistical uncertainties are given for σ_R^2. The claim that deviations from k^3 scaling at large β are 'likely attributable to the limited number of bubbles' is not supported by any convergence test or error estimate. Since Eq. (20) is fitted to these same σ_R^2 values, the fitting coefficients A=0.0038, a=0.043, b=1.77 inherit these unquantified uncertainties. The observational constraints in Sec. V and Table I are computed using this template, so the presented exclusion regions are not accompanied by a systematic error budget. Please provide error bars (e.g., jackknife or bootstrap over histories) and a convergence test in the number of bubbles/realizations, or state clearly that the template is a central-value fit.","section":"Sec. IV.B; Fig. 2; Eq. (20)"},{"comment":"The fitting parameters in Eq. (20) are fitted to the same DeltaPT simulations that produce the central quantities, and the same template is then used to derive all observational constraints and the SIGW spectra in Sec. V. This is not circular for the direct PBH conclusion, which is computed from the simulated R distribution, but it limits the independent validation of the template. The manuscript would be stronger if the simulation input (or at least the σ_R^2 data points) were provided as supplementary material, and if the template fit were validated against out-of-sample parameter points. Without this, the Table I constraints should be seen as dependent on an unvalidated functional form.","section":"Sec. V; Eq. (20); Table I"}],"minor_comments":[{"comment":"The horizontal axis label in the right panel appears garbled: '1016 × 100 2 × 101' should read '10^0, 2×10^1' (or similar). Please fix this and ensure all axis labels are clear.","section":"Sec. IV.B; Fig. 2"},{"comment":"The text states 'For each super-horizon wave number k > kmax', but the figures and earlier discussion concern modes k < k_max that re-enter after percolation. Please correct this apparent typo.","section":"Sec. IV.A"},{"comment":"The cancellation of the divergence in S_Vr when ρ̇_r = 0 is described only in words. Since the prefactor in Eq. (17) also contains 1/(6H ρ̇), it would be helpful to show the explicit cancellation or state the limiting behavior when both ρ̇_r and ρ̇_V approach zero.","section":"Eq. (18) and surrounding text"},{"comment":"Typo: 'heaviside' should be 'Heaviside'. Also, the phrase 'Θ(k_max − k)' should be written with a space or consistent notation.","section":"Eq. (20)"},{"comment":"The ranges for β/H_n (e.g., '>~ 35−80') are not explained. Are these ranges due to different constraints inside a column, or to the adopted reference values? Please clarify in the text or caption.","section":"Table I"}],"recommendation":"major_revision","confidential_remarks":"The paper reaches a conclusion similar to Ref. [20] (Franciolini, Gouttenoire, Jinno), and the authors do cite it and state agreement. However, the added value of the present paper—the fitting formula, the observational constraints, and the PTArcade analysis—depends on the same linearization assumption as the core PBH suppression. The editor may wish to ask the authors to make the DeltaPT input data or code available, since the reproducibility of the central suppression claim is otherwise limited. The paper is not internally inconsistent, but the load-bearing linearization issue and missing error bars require a revision before I would recommend acceptance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this paper confirms the Franciolini-Gouttenoire-Jinno result that separate-universe density contrasts from slow, strongly supercooled FOPTs do not translate into PBH-forming curvature perturbations once gauge ambiguities are handled. The genuinely new items are the empirical power-spectrum template in Eq. (20) and the NANOGrav-15 posterior analysis showing SIGWs from these perturbations are subdominant. Both are worth having.\n\nThe formalism is standard multi-fluid perturbation theory, taken from Malik-Wands et al., and the paper applies it carefully. The explicit statement that the result agrees with Ref. [20] is honest, and the agreement between two different routes to the suppression is the main reason I trust the qualitative conclusion. The fitting formula has the right asymptotics (k^3, beta^-5) and gives phenomenologists a ready-to-use template. The PTA analysis is a useful addition; the conclusion that the secondary SIGW component barely shifts posteriors is clean.\n\nThe soft spot is the one the stress-test note flags. In Sec. IV.A the authors assume delta_NG from DeltaPT can be approximated, at linear order, by density contrasts on a specific but unknown time slicing. The unknown slicing does cancel in the relative entropy perturbation S_Vr at linear order, as they effectively rely on. What is not protected is the linearization itself. The paper never shows delta_NG << 1; Fig. 1 plots values that reach O(1), and no convergence test in delta is reported. If the typical delta is O(0.1-0.4), second-order corrections to the gauge transformation and to the finite-patch-to-global mapping could enter at a level comparable to the tiny |R| values (~1e-4 or less) that drive the suppression. So the central PBH-unlikely claim is plausible but not yet robust against this failure mode. A nonlinear separate-universe zeta check would settle it.\n\nOther soft spots are minor by comparison. The fitting coefficients in Eq. (20) come from the same DeltaPT runs used to derive the underlying sigma_R^2, so the template is not an independent prediction; that is fine, but it should be stated more clearly. There are no error bars on sigma_R^2 or the delta distributions. The code and simulation outputs are not shipped, so the pipeline is not reproducible from the paper alone. The paper itself flags the monochromatic-spectrum caveat for the UCMH constraints, which is good.\n\nWho should read it: anyone working on PBH formation from FOPTs or on FOPT gravitational-wave backgrounds. It will be cited. I'd send it to a serious referee; a good referee will ask for the linearization check and perhaps the code. It is not a desk reject.","headline":"Solid confirmation that slow-FOPT super-horizon inhomogeneities don't form PBHs once gauge issues are handled; the new fitting template and PTA analysis are useful, but the linearization assumption needs a check.","tokens_in":13126,"tokens_out":2949,"would_cite":true,"duration_ms":31433,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A gauge-invariant calculation shows that curvature perturbations from slow first-order phase transitions are far smaller than earlier separate-universe estimates, making primordial black hole formation from this mechanism very unlikely.","keywords":["first-order phase transitions","curvature perturbations","primordial black holes","gauge invariance","separate universe","supercooled transitions","scalar-induced gravitational waves","multi-fluid perturbation theory"],"falsifier":"A fully nonlinear simulation of bubble nucleation in an expanding universe that computes the comoving curvature perturbation without invoking the linear-slicing assumption would settle the issue; if it produced a density contrast distribution with a tail above about 0.45 for some parameters, the central suppression claim would be refuted.","tokens_in":12194,"feed_emoji":"🌌","tokens_out":3100,"duration_ms":33771,"temperature":0.7,"pith_summary":"The paper argues that the super-horizon density contrasts produced when a slow, strongly supercooled cosmological first-order phase transition nucleates bubbles are much smaller than previously thought. Working with a two-fluid system of radiation and vacuum, the authors compute the gauge-invariant comoving curvature perturbation and find its amplitude is suppressed relative to earlier separate-universe results. Consequently, the density contrast distribution lies below the typical black-hole formation threshold, so primordial black holes are unlikely to form through this mechanism. The paper also provides a fitting formula for the curvature power spectrum and shows that scalar-induced gravitational waves from these perturbations are subdominant.","feed_headline":"Slow phase transitions can't make black holes","feed_subtitle":"Gauge-invariant calculation cuts earlier estimates, leaving density contrasts below the collapse threshold.","key_machinery":"The central object is the gauge-invariant comoving curvature perturbation R (equivalently zeta), evolved on super-Hubble scales via d(zeta)/dt = -H/(rho+P) deltaP_nad, where deltaP_nad is the total non-adiabatic pressure perturbation. For a two-fluid system of radiation and vacuum, the intrinsic non-adiabatic pressure vanishes because each fluid has a definite barotropic equation of state, leaving only the relative entropy perturbation between the fluids. The authors avoid gauge ambiguities by treating the density contrasts from separate-universe-style nucleation simulations as linear-order perturbations on a specific but unspecified time slicing, then use the gauge-invariant equations to co","core_discovery":"The central claim is that the comoving curvature perturbation generated by super-horizon inhomogeneities from slow first-order phase transitions is much smaller than implied by previous separate-universe studies, and the resulting density contrast does not exceed the conventional primordial black hole formation threshold. Using gauge-invariant multi-fluid perturbation theory, the authors evolve the curvature perturbation on super-Hubble scales, accounting for the relative entropy perturbation between radiation and vacuum. They derive a fitting formula for the curvature power spectrum that scales as k^3 at small k and beta^-5 at large beta, but with an amplitude suppressed compared to earlier","pith_inferences":["If the suppression holds, the window for primordial black hole dark matter from this specific mechanism closes, though other formation channels such as false vacuum islands remain unaffected.","The linear-slicing assumption could be tested by comparing against a fully nonlinear simulation of bubble nucleation; if nonlinear corrections are large, the suppression may change.","The fitting formula offers a ready way to forecast LISA-era constraints on supercooled transitions, which the paper only sketches via existing datasets.","The result underscores that any claim of black hole production from phase transitions must specify the gauge in which the density contrast is defined."],"forward_implications":["Primordial black hole formation from super-horizon inhomogeneities at slow first-order phase transitions is strongly disfavored; the density contrast distribution stays below the critical threshold.","Scalar-induced gravitational waves from these curvature perturbations are subdominant compared to primary bubble-collision or sound-wave signals.","The derived fitting formula for the curvature power spectrum can be used to translate current and future constraints on primordial perturbations into constraints on phase transition parameters.","Future radio and dark-matter-halo observations may constrain both slow and fast phase transitions, with the suppressed spectrum shifting the reachable parameter space.","The gauge-invariant treatment provides a template for evaluating other cosmological observables arising from large-scale phase-transition inhomogeneities."],"fun_headline_variants":["Gauge-invariant math rules out black holes from slow transitions","Gauge-invariant analysis damps phase-transition perturbations","Curvature perturbations from slow phase transitions: smaller than thought","No black holes from slow phase transitions, gauge-invariant says","Gauge-invariant analysis shrinks phase-transition curvature"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The calculation assumes that the density contrasts from separate-universe simulations can be approximated, at linear order, as perturbations on a specific but unspecified time slicing, so that the gauge-invariant multi-fluid formalism applies.","fun_headline_variants_meta":{"raw":{"variants":["Gauge-invariant math rules out black holes from slow transitions","Gauge-invariant analysis damps phase-transition perturbations","Curvature perturbations from slow phase transitions: smaller than thought","No black holes from slow phase transitions, gauge-invariant says","Gauge-invariant analysis shrinks phase-transition curvature"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000737,"raw_usage":{"total_tokens":3053,"prompt_tokens":594,"completion_tokens":2459,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":338,"completion_tokens_details":{"reasoning_tokens":2375}},"tokens_in":338,"tokens_out":2459,"duration_ms":17347,"temperature":1.0,"reasoning_tokens":2375,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T09:12:07.163330+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A fully nonlinear simulation of bubble nucleation in an expanding universe that computes the comoving curvature perturbation without invoking the linear-slicing assumption would settle the issue; if it produced a density contrast distribution with a tail above about 0.45 for some parameters, the central suppression claim would be refuted.","supporting_citations":[],"review_version":1}