{"id":"c9792daf-d7a5-4b94-ba62-fced6b8aad7c","arxiv_id":"2501.13691","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The paper argues that discarding non-light-speed modes of the second-order tensor perturbation makes the induced gravitational wave energy density gauge-invariant in both adiabatic and isocurvature scenarios.","lead":"This paper proposes a filtering rule, based on the Sommerfeld radiation condition, for separating real gravitational waves from spurious non-propagating modes in second-order tensor perturbations of the early universe. If the rule is right, the predicted energy density of scalar-induced gravitational waves is the same in every gauge, fixing a long-standing inconsistency in cosmological computations.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Applying the Sommerfeld filter to the fixed-(u,v) kernel before the momentum integral is the load-bearing step: the paper does not prove that the discarded I_chi cannot contribute to the observable h_k/Omega_GW, and it leaves the canonical-observer link as a conjecture.","rationale":"The reader's weakest assumption correctly identifies the load-bearing premise: the Sommerfeld criterion is applied to the kernel before the momentum integral and equated with what a canonical observer measures, with the equivalence left conjectural. My reading of the paper confirms this is not a peripheral technicality but the very step that converts the gauge-dependent tensor perturbation into a gauge-invariant observable. The mathematical steps after the filter are coherent and the analytical examples are consistent, so the verdict CONDITIONAL is appropriate. I found no independent fatal flaw; the strongest remaining need is a proof that the filtered part is exactly the radiative part of the momentum-integrated h_k, or an explicit derivation of the canonical-observer correspondence. The concrete test I propose would settle this by checking whether I_chi survives the momentum integral as a light-speed radiative tail, and by comparing against a manifestly gauge-invariant observable such as Psi_4. If the test supports the paper's filtering, the gauge-invariance claim would be considerably strengthened; if not, the central claim would need revision.","tokens_in":11223,"tokens_out":18363,"duration_ms":180614,"concrete_test":"Compute, for the comoving gauge, the full momentum integral in Eq. (2.2) with I = I^N_AD + I_chi of Eq. (3.3) for a sharply peaked spectrum, and extract the large-x component oscillating as sin x or cos x from h_k. Then compare with the filtered Newtonian kernel. If the momentum-integrated I_chi yields no light-speed radiative tail (e.g., decays faster than 1/x or oscillates at x/sqrt(3)), the filter's mode-by-mode application is vindicated; if it does yield a 1/x sin x term, the central claim fails. A complementary decisive check is to compute the radiative part via the Weyl scalar Psi_4 on the canonical observer tetrad of Ref. [27] for the same gauge and compare with the filtered Omega_GW.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that after discarding I_chi and Delta I, the remaining I_{Y,J} gives the physical Omega_GW in every gauge. This rests on the assertion that I_chi contains no light-speed (sin x or cos x) component, so it violates the Sommerfeld condition. But the Sommerfeld condition, Eq. (3.1), is a statement about the real-space field h_ij (or the momentum-integrated h_k), not about the integrand I(u,v,x) at fixed u,v. The observable h_k from Eq. (2.2) involves an integral over d^3p with the projection e_ij p^i p^j and the power spectrum; sub-light-speed oscillations in I_chi can, after this integral, produce contributions whose phase and decay differ from those of a single term. The paper's three bullets argue absence of sin x/cos x in I_chi for fixed u,v, but do not evaluate the momentum integral. Moreover, the claim \"regardless of the choice of alpha and L\" is stronger than what is shown: if a gauge vector has light-speed oscillatory components, products T_Y(ux)T_Y(vx) contain cos x on the surface u+v=1. For the standard gauges the filtering is plausible, but the paper itself states that the equivalence with a canonical observer is a conjecture. Hence the gauge-invariance theorem is conditional on a prescription whose physical content is not derived.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a boundary-condition-based filtering method to define the physical part of second-order scalar-induced tensor perturbations (SIGWs). In Newtonian gauge the kernel I_X is decomposed into two Bessel-function components I_X,Y and I_X,J plus a remainder Delta I_X that vanishes in the sub-horizon limit. An arbitrary gauge transformation adds an extra piece I_chi built from products of first-order transfer functions. The authors argue that I_chi contains no light-speed sin x or cos x terms at fixed (u,v), violates the Sommerfeld condition, and should be discarded. They conclude that the filtered Omega_GW is finite and gauge-invariant for both adiabatic and isocurvature perturbations, and they support this with numerical kernel plots in a variety of gauges.","tokens_in":11536,"tokens_out":6305,"duration_ms":63862,"significance":"If fully established, the result would provide a simple and elegant resolution of a long-standing problem in cosmological perturbation theory. The paper is clearly organized, gives explicit formulas for the gauge-transformation kernel I_chi in the comoving and synchronous examples, and shows numerically that the filtered kernels collapse to a common ~1/x behavior in several gauges. The proposed criterion is physically motivated and is not fitted to data. However, the central claim is conditional on a prescription whose physical identification is explicitly left as a conjecture, and the filtering is applied at the level of the fixed-(u,v) kernel rather than to the observable momentum-integrated field. Those gaps are load-bearing rather than cosmetic.","major_comments":[{"comment":"The Sommerfeld condition is stated for the field h_ij (Eq. (3.1)), but the filtering is imposed on the fixed-(u,v) kernel I_X(u,v,x) before the momentum integral in Eq. (2.2). A kernel term that contains only sound-speed oscillations at fixed u,v can, in principle, contribute to h_k after integration over d^3p with the projection e_ij p_i p_j and the power spectrum; the late-time phase and decay of the integrated result need not coincide with those of any single term in the integrand. The three bullets in Sec. 3 demonstrate only the pointwise absence of sin x and cos x in I_chi; they do not evaluate the momentum integral. Since this is exactly the step that removes the x^2 and x^4 divergences, the central claim requires either an explicit estimate of the integrated I_chi contribution to Omega_GW or a proof that the momentum integration cannot convert sound-speed kernel oscillations into a light-speed gravitational-wave component.","section":"Sec. 3, bullet list after Eq. (2.10)"},{"comment":"The statement that I_chi 'definitely violates the Sommerfeld boundary condition ... regardless of the choice of w, alpha and L' is stronger than what is shown. The argument uses only the generic functional form T_Y(ux)T_Y(vx). If a gauge parameter has a light-speed oscillatory transfer function, for example T_alpha(ux)=sin(ux), then sin(ux)sin(vx) contains cos((u+v)x), which has a light-speed component on the surface u+v=1. The same resonance mechanism that produces sin x in the standard Newtonian-gauge kernel can therefore operate in I_chi for certain gauge choices. The claim is plausible for the standard gauges studied in the paper, where T_alpha and T_L inherit sound-speed oscillations, but it is not proven for arbitrary alpha and L.","section":"Sec. 4 and Sec. 3"},{"comment":"The identification of the Sommerfeld-satisfying part of h_ij with what 'a canonical observer' measures is left as a conjecture: the paper states that a rigorous proof of this correspondence is an interesting direction for future research. Because the filtered Omega_GW is gauge-invariant by construction once I_chi is discarded, the physical content of the result rests entirely on this identification. Without an independent tetrad/observer derivation or an operational definition of the filtered field, the paper establishes conditional gauge invariance of a prescribed projection, not gauge invariance of the observable GW energy density as directly measured or defined from the metric.","section":"Sec. 3, Figs. 1 and 2"},{"comment":"The numerical demonstration in Figs. 1 and 2 plots the kernel at fixed u=v=1, not the momentum-integrated h_k or Omega_GW. The collapse of the fixed-kernel curves after filtering is suggestive but is not by itself evidence that the filtered spectrum is identical to the Newtonian-gauge spectrum after the momentum integral over all u and v.","section":"Sec. 3, Figs. 1 and 2"}],"minor_comments":[{"comment":"Eqs. (B.3) and (B.4) are split into two display equations in a confusing way; they should be presented as a single equation for I_ISO(u,v,x).","section":"Appendix B"},{"comment":"Eq. (A.2) contains an unbalanced parenthesis in the first line, and the phrase 'vanishes' in Eq. (2.8) should be 'vanish'.","section":"Appendix A"},{"comment":"The captions of Figs. 1 and 2 should state explicitly which curves correspond to which gauge and should define the 'after filtering' procedure; currently they are not self-contained.","section":"Sec. 3"},{"comment":"The remark about using h'_ij(k)=k h_ij(k) being paradoxical for non-light-speed components is important and deserves a more formal statement, for example a definition of when the replacement is allowed after filtering.","section":"Sec. 4"},{"comment":"There are several typos, including 'genearl', 'obatin', and 'Iχ' mixed with 'I_chi'; a careful proofreading pass is needed.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The paper's central assertion is stronger than what is demonstrated: the filtering prescription is applied to the fixed-(u,v) kernel, and the canonical-observer interpretation is explicitly conjectural. I would ask the authors to either provide a momentum-integral-level argument that I_chi has no physical GW component, or to reframe the paper as a proposal for a gauge-invariant definition of the SIGW energy density whose physical interpretation remains to be established. The paper would then be a valuable contribution rather than a complete resolution of the gauge problem."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This paper offers a concrete fix for the gauge-dependence problem in scalar-induced gravitational waves: apply a Sommerfeld outgoing-wave filter to the second-order tensor kernel, discard the non-propagating pieces, and compute Omega_GW from what remains. The filtered spectrum is finite and gauge-invariant for both adiabatic and isocurvature perturbations, and the authors provide explicit filtered kernels in Appendix B that others can use. The algebra is coherent and the diagrams showing divergences removed are convincing. That is a real step forward.\n\nThe genuinely new content is the filtering prescription itself: separate the kernel into physical oscillating components (Y/J Bessel terms) and unphysical parts (Delta I_N and I_chi), then discard the latter because they oscillate at the sound speed rather than the speed of light. The decomposition idea is credited to Inomata & Terada [20] and the canonical-observer language to Cai et al. [27], but the authors work it out for arbitrary gauges and for isocurvature modes, with explicit formulas. The extension to isocurvature is useful and the appendix formulas are reproducible.\n\nThe load-bearing step is applying the Sommerfeld condition to the fixed-(u,v) kernel I(u,v,x) before the momentum integral in Eq. (2.2). The paper argues that I_chi contains no sin x or cos x at fixed u,v, but the observable h_k is an integral over u,v weighted by the power spectrum. Sub-light-speed oscillations in I_chi could, in principle, leave a light-speed tail after integration. I suspect the tail vanishes because the phase speeds are bounded by the sound speed, but the paper does not prove it. The authors themselves state in Section 4 that the connection to a canonical observer is a conjecture. So the gauge-invariance theorem is conditional: it holds if the Sommerfeld-filtered kernel is exactly what a physical observer measures, and that is precisely the part left open.\n\nThere is also a mild circularity: the physical part is defined as the part that satisfies the Sommerfeld condition, so discarding I_chi guarantees gauge invariance by construction. The criterion is independent and physically motivated, so this is not disqualifying, but it means the paper establishes an internal consistency result plus a conjecture about observability, not a fully derived physical statement.\n\nThis deserves a serious referee. The calculations are explicit and reproducible, the gap is clearly flagged by the authors, and the filtered kernels in Appendix B will be used regardless. A referee should push for a proof that the momentum integral of I_chi has no 1/x sin x/cos x tail and for a concrete observer model. I would not cite this as a resolution of the gauge problem until that is done, but I would cite the filtered kernels for practical spectrum computations.","headline":"A concrete Sommerfeld-filtering prescription that makes Omega_GW gauge-invariant and finite, but the link to what a physical observer measures is left as an explicit conjecture.","tokens_in":12026,"tokens_out":3335,"would_cite":true,"duration_ms":30417,"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":"This paper claims that applying the Sommerfeld outgoing-wave boundary condition to the Fourier-space kernel of scalar-induced gravitational waves filters out gauge-dependent non-luminal modes, making the energy density $\\Omega_{\\rm GW}$…","keywords":["scalar-induced gravitational waves","gauge invariance","Sommerfeld boundary condition","cosmological perturbation theory","adiabatic perturbations","isocurvature perturbations","gravitational wave energy density","radiation domination"],"falsifier":"A concrete check: search the gauge-transformation kernel $I_\\chi(u,v,x)$ in Eq. (2.10) for a term whose large-$x$ behavior is $x^{-\\beta-1/2}\\sin x$ or $x^{-\\beta-1/2}\\cos x$; if such a term exists for some $(u,v)$, it satisfies the Sommerfeld condition and would survive the filter, giving a gauge-dependent $\\Omega_{\\rm GW}$. Alternatively, compute the Newman-Penrose Weyl-scalar energy in comoving gauge and compare it with the filtered $\\Omega_{\\rm GW}$; a discrepancy would show that the filter does not correspond to a physical measurement.","tokens_in":10992,"feed_emoji":"🌊","tokens_out":7791,"duration_ms":65917,"temperature":0.7,"pith_summary":"The paper takes on a long-standing problem in cosmological perturbation theory: the predicted energy density of scalar-induced gravitational waves (SIGWs) depends on the gauge used to compute it, and in some gauges it diverges. It proposes that the second-order tensor perturbation $h_{ij}$ is not purely gravitational radiation; it mixes in non-propagating, gauge-dependent 'virtual' modes. The proposed fix is to apply the Sommerfeld outgoing-wave boundary condition to each Fourier-space kernel mode and keep only components that oscillate at the speed of light and decay as outgoing spherical waves. After this filtering, the energy density $\\Omega_{\\rm GW}$ becomes gauge-invariant and finite in arbitrary gauges, for both adiabatic and isocurvature perturbations. A sympathetic reader would care because it supplies a physical criterion for what counts as a gravitational wave in an expanding universe and makes SIGW predictions used to probe small-scale primordial physics independent of gauge choice.","feed_headline":"Sommerfeld filter makes induced GW spectra gauge invariant","feed_subtitle":"Non-luminal virtual modes are filtered out, so Ω_GW agrees in every gauge for adiabatic and isocurvature perturbations.","key_machinery":"The machinery is a mode-by-mode filter applied to the kernel function $I_X(u,v,x)$ entering the Fourier-space solution for $h^\\lambda_k$. The kernel is split as $I^N_X = I^N_{X,Y} + I^N_{X,J} + \\Delta I^N_X$; the first two terms describe physical radiation because they asymptote to outward-propagating light-speed waves, $x^{-\\beta-1/2}\\sin x$ and $x^{-\\beta-1/2}\\cos x$. The filter is the Sommerfeld condition $\\lim_{t\\to\\infty}(\\partial_\\eta + \\partial_r)(a h_{ij}) = 0$, applied at the level of $I(u,v,x)$ before the momentum integrals in Eq. (2.3). Components that oscillate at the sound speed, such as $\\sin(ux/\\sqrt{3})$ in radiation domination, or that contain no Green's-function propagation factor and hence no wave-propagation effect, are discarded as non-physical.","core_discovery":"The central claim, stated for arbitrary gauges, is that the physical $\\Omega_{\\rm GW}$ is always contributed by the first two oscillating components of the kernel in Eq. (2.4), the terms behaving as $x^{-\\beta-1/2}\\sin x$ and $x^{-\\beta-1/2}\\cos x$, and that this result equals the Newtonian-gauge result. The remaining part of the kernel—the $\\Delta I^N_X$ piece even in Newtonian gauge, and the $I_\\chi$ piece generated by any gauge transformation—violates the Sommerfeld boundary condition because it oscillates at the sound speed (e.g., $\\sin(ux/\\sqrt{3})$) and does not propagate at the speed of light. These pieces are declared non-physical radiation and are filtered out. The paper shows this concretely for radiation domination: the previously reported $x^2$ divergence in comoving gauge and the $x^4$ divergence in synchronous gauge disappear after filtering, for adiabatic and isocurvature sources respectively. The same filtered kernel is obtained in every gauge, so the late-time GW spectrum becomes unique.","pith_inferences":["If accepted, the same boundary-condition criterion should apply to other local gravitational-wave observables such as geodesic deviation or Newman-Penrose scalars, so the filtered $\\Omega_{\\rm GW}$ should match a tetrad measurement; the paper leaves this correspondence as a conjecture.","The method suggests a sharp conceptual distinction between tensor perturbations and gravitational radiation: only the wave-zone, luminal, decaying part carries energy, so GW energy densities should be defined after a wave-zone projection rather than by averaging all quadratic tensor modes.","A natural stress test is to apply the same filtering to SIGWs with a general sound speed $c_s \\neq 1/\\sqrt{3}$; the filter's reliance on light-speed propagation may need modification, and gauge invariance could fail if source transfer functions contain luminal oscillations.","The decomposition into physical and virtual modes aligns with a quantum-field-theoretic picture: the filtered part is on-shell graviton radiation, while the discarded part is off-shell virtual exchange; this connection is an editorial extension not proven in the paper."],"forward_implications":["SIGW spectra for both adiabatic and isocurvature perturbations can be computed in any gauge without gauge-suitability arguments; the answer always matches the Newtonian-gauge result.","The divergent $\\Omega_{\\rm GW}$ scalings reported in comoving, synchronous, uniform-density, and other gauges are explained as contamination from non-luminal, non-propagating modes rather than physical divergences.","Observational predictions for pulsar-timing-array and future space-based gravitational-wave backgrounds from scalar-induced sources become gauge-robust, since the filtered spectra are uniquely defined.","The analytic filtered kernel formulas in Appendix B give ready-to-use $\\Omega_{\\rm GW}$ spectra in radiation domination for both adiabatic and isocurvature sources.","The filtering resolves the internal inconsistency of using $h'_{ij}(k) = k h_{ij}(k)$ to compute energy density in gauges where part of $h_{ij}$ does not propagate at light speed."],"supporting_citations":[{"why":"First pointed out that the second-order tensor mode splits into a propagating gravitational-wave part and a non-GW part, the idea the filtering method builds on.","marker":"[20]"},{"why":"Derives the gauge-transformation kernel $I_\\chi$ and the comoving-gauge $x^2$ divergence that the filtering must remove.","marker":"[22]"},{"why":"Shows the $x^4$ gauge-dependent divergence for isocurvature-induced SIGWs in synchronous gauge, the target case for the unified claim.","marker":"[24]"},{"why":"Introduces canonical observer tetrads for defining gravitational-wave energy, which the paper conjectures implements the same filtering.","marker":"[27]"},{"why":"Supplies the semianalytic Newtonian-gauge kernel $I_{\\rm AD}$ used to identify the physical oscillating components during radiation domination.","marker":"[29]"},{"why":"Provides the general-background transfer functions for adiabatic perturbations used to argue $\\Delta I_{\\rm AD}$ vanishes in the sub-horizon limit.","marker":"[30]"},{"why":"Provides the isocurvature transfer functions and kernel expressions needed for the isocurvature case.","marker":"[31]"},{"why":"States the Sommerfeld outgoing-wave boundary condition for gravitational waves around flat spacetime that the filter is based on.","marker":"[33]"}],"fun_headline_variants":["Filtering non-luminal modes makes induced GW gauge invariant","Sommerfeld criterion cures gauge dependence of induced GWs","Physical GW spectrum unique after Sommerfeld filter","Gauge-invariant induced GWs via boundary condition filter"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument depends on the assumption that the flat-spacetime outgoing-wave condition, applied to each Fourier-space kernel mode before integrating over momenta, picks out exactly the part of the tensor perturbation that a physical observer would measure as gravitational radiation in an expanding universe.","fun_headline_variants_meta":{"raw":{"variants":["Filtering non-luminal modes makes induced GW gauge invariant","Sommerfeld criterion cures gauge dependence of induced GWs","Physical GW spectrum unique after Sommerfeld filter","Gauge-invariant induced GWs via boundary condition filter"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000211,"raw_usage":{"total_tokens":1377,"prompt_tokens":870,"completion_tokens":507,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":486,"completion_tokens_details":{"reasoning_tokens":441}},"tokens_in":486,"tokens_out":507,"duration_ms":4519,"temperature":1.0,"reasoning_tokens":441,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T15:42:13.365684+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A concrete check: search the gauge-transformation kernel $I_\\chi(u,v,x)$ in Eq. (2.10) for a term whose large-$x$ behavior is $x^{-\\beta-1/2}\\sin x$ or $x^{-\\beta-1/2}\\cos x$; if such a term exists for some $(u,v)$, it satisfies the Sommerfeld condition and would survive the filter, giving a gauge-dependent $\\Omega_{\\rm GW}$. Alternatively, compute the Newman-Penrose Weyl-scalar energy in comoving gauge and compare it with the filtered $\\Omega_{\\rm GW}$; a discrepancy would show that the filter does not correspond to a physical measurement.","supporting_citations":[{"cited_title":"On the energy of gravitational waves","cited_arxiv_id":"2109.06864","evidence_quote":"Introduces canonical observer tetrads for defining gravitational-wave energy, which the paper conjectures implements the same filtering."}],"review_version":1}