{"id":"4b344327-ad05-40d2-a731-b065756cb04e","arxiv_id":"2506.20733","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Exact-in-velocity formulas for the O(G^3) nonlinear gravitational memory multipoles from two-body scattering, derived with scattering amplitudes and reverse unitarity, and matched to post-Newtonian results.","lead":"The paper computes, for the first time and for any relative speed, the permanent 'memory' shift imprinted on gravitational waves by the gravity waves themselves when two compact objects scatter. It provides exact velocity-dependent formulas for the leading nonlinear memory multipoles and checks them against earlier slow-motion approximations.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The exact-in-velocity evaluation of δF^{ℓm} (Eqs. (15)-(24), Table I) rests on an unshown IBP reduction and master-integral solution deferred to [82]; the small-v checks do not test the finite-σ functions, so the headline coefficients are not independently checkable from this preprint.","rationale":"The paper's reason for being is Eq. (23): first exact-in-velocity O(G^3) nonlinear memory multipoles. Everything needed to reach that equation is either standard (Weinberg soft theorem, reverse unitarity, STF multipole decomposition) or explicitly checked (small-velocity agreement, agreement with Wiseman-Will and Favata). The one step that is both essential and not independently verifiable from the preprint is the IBP reduction and master-integral solution that turn the phase-space integral (20) into the closed forms in Table I and Appendix III. The reader's weakest assumption identifies exactly this step, and I agree with that identification. The low-velocity checks are genuinely valuable evidence, but they constrain only the σ→1 expansion; a subtle error in the finite-σ master integrals or boundary conditions could preserve the first many Taylor coefficients while changing the exact functions. The coherent-state conclusion in the final paragraphs is also stronger than the computation strictly demonstrates, but it is not load-bearing for the numerical content of Eq. (23), so I do not make it the primary concern. The appropriate verdict remains CONDITIONAL: the central claim is plausible and well supported in outline, but full acceptance should wait for the detailed reduction in [82] or an independent reproduction. Hence I recommend no change to the reader's verdict.","tokens_in":16244,"tokens_out":10687,"duration_ms":132019,"concrete_test":"Run the IBP reduction of the ℓ=2, m=2 integral in Eq. (20) with LiteRed or FIRE, express it in the master-integral basis stated after Eq. (20), and independently solve the soft-region master integrals by the method of canonical differential equations with boundary conditions at σ=1 fixed by the known v→0 expansion; then compare the resulting f22_1, f22_2, f22_3 against Table I. Exact agreement would settle the concern; a mismatch would localize the erroneous reduction or boundary condition.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim, Eq. (23), gives exact-in-σ polynomials f^{ℓm}_i. The route to these polynomials is Eq. (20) reduced by IBP to master integrals, solved by canonical differential equations in the soft region, with the text after Eq. (23) deferring details to [82]. The IBP reduction and, especially, the boundary conditions at σ=1 are not shown. An error there would change every entry in Table I and Appendix III. The provided checks are real but cannot rule this out: the agreement with Ref. [87] tests the Taylor expansion around σ=1 through v^19 (9PN), and the agreement with [88,89] tests only the first nonzero term; neither fixes the exact finite-σ functions. The soft-theorem consistency check of Eq. (14) validates the amplitude representation of δF, not the evaluation of the phase-space/master integrals. The same caveat applies to the legibility of Eqs. (12)-(14), whose diagrams are not fully visible in the supplied text, so the two-loop soft-limit input cannot be inspected directly. This is a verifiability concern, not a claimed error; the route is standard and the low-velocity limits are encouraging.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript presents a computation of the nonlinear gravitational memory multipoles at O(G^3) in the post-Minkowskian expansion for the scattering of two compact objects, keeping the relative velocity arbitrary. The authors use the scattering-amplitudes representation of the waveform, a multipolar decomposition, and the reverse-unitarity method to derive the master formula (23), with explicit polynomials f_i^{ℓm}(σ) given in Table I and Appendix III. The results are checked against the small-velocity expansion of a companion calculation [87] through 9PN and against earlier leading-order results [88,89]. The central claim is that this provides, for the first time, the exact-in-velocity nonlinear memory at O(G^3), completing the gauge-invariant non-analytic-in-frequency part of the multipolar waveform at this order.","tokens_in":16440,"tokens_out":4304,"duration_ms":50836,"significance":"If the result is correct, it is a significant step: it is the first exact-in-velocity determination of the nonlinear memory at O(G^3), and it provides a concrete benchmark against which future waveform calculations—particularly those based on worldline effective field theory, self-force, or numerical relativity—can be tested. The paper also connects the nonlinear memory to the low-frequency limit of a two-loop amplitude, strengthening the amplitude/classical-GR dictionary. The inclusion of the master formula, the accompanying polynomials up to ℓ=5, and the machine-readable ancillary file nlmMultipoles.m are valuable assets. The cross-checks against small-velocity results and the consistency with the Weinberg soft theorem are nontrivial and are appropriately highlighted. However, the exact-in-velocity evaluation itself is not fully exhibited in this preprint, which limits the degree to which the headline coefficients can be independently verified.","major_comments":[{"comment":"The central exact-in-velocity result rests on an IBP reduction to master integrals and their evaluation via canonical differential equations in the soft region, but none of these steps are shown. The text states after Eq. (23) that 'more details will be provided in [82]'. Consequently, the polynomials f_i^{ℓm}(σ) in Table I and Appendix III cannot be independently verified from this preprint alone. The small-velocity checks against Ref. [87] (through 9PN) and Refs. [88,89] (leading order) test only the Taylor expansion around σ=1 and do not fix the finite-σ log((σ+1)/2) and arccoshσ terms that appear in Eq. (24). Because an error in the master-integral boundary conditions or in the IBP reduction would change every entry in Table I and Appendix III, this is a load-bearing gap. The authors should either include the master-integral definitions, the IBP reduction master decomposition, and the boundary conditions, or provide a detailed ancillary derivation file, before the exact-in-velocity claim can be accepted as fully supported.","section":"Eq. (14)"},{"comment":"The two-loop amplitude representation in Eq. (14) is the key input that identifies the nonlinear-memory contribution, but the diagrammatic content is not visible in the supplied text: the figures are missing or not rendered. Since the identification of which terms contribute to δF at leading order in ω depends on the detailed structure of these diagrams, the reader cannot inspect whether all relevant contributions (including the various cuts in the round parentheses) are accounted for. The authors should provide a legible version of these diagrams or, alternatively, a textual specification of the diagram topologies and their numerator factors. This is essential for the reproducibility of the central decomposition.","section":"Conclusions"},{"comment":"The concluding paragraph makes a strong physical claim: that the way the nonlinear memory arises from the soft limit of a 2→4 amplitude shows that 'classical gravitational radiation is not completely described by a (non-squeezed) coherent state', which is stated without a derivation or a detailed argument in the manuscript. This goes beyond the computation presented here and would require an explicit analysis of the state or of the higher-order correlators. The authors should either provide that argument or soften the claim to a conjecture/future work.","section":"Checks"},{"comment":"The validation against Ref. [87] is a comparison with a companion calculation by overlapping authorship, and both the present computation and Ref. [87] rely on the same scattering-amplitudes framework and the same expression for ρ(k). The agreement is a strong consistency check, but it is not a fully independent verification of the exact-in-velocity functions. The authors are encouraged to add an additional test that is sensitive to the finite-σ structure, for example a numerical evaluation of the phase-space integrals in Eqs. (15)-(20) at representative values of σ, or a comparison with an independent analytic method in the ultra-relativistic limit σ→∞.","section":"Summary"}],"minor_comments":[{"comment":"The phrase 'the nonlinear memory f^{μν} in (3)' appears to be a typo: f^{μν} in Eq. (4) is the linear memory, while the nonlinear memory is δF^{μν}. Please correct this to 'linear memory' or 'memory'.","section":"Eq. (18)"},{"comment":"In Eq. (18), the spin-weighted spherical harmonics Y^{ℓm}_{±2} are used for the multipole decomposition, but the notation Y^{ℓm*}_{±2} is not defined explicitly in the main text; the definition is deferred to Appendix I. This is acceptable, but a brief reminder in the main text would improve readability.","section":"Eq. (24)"},{"comment":"The functional form in Eq. (24) contains log((σ+1)/2) and arccoshσ/√(σ²−1); it would be helpful to state explicitly that these combinations are regular at σ=1 and to comment on their behavior as σ→∞, since these are the limits used in the checks.","section":"Ancillary file"},{"comment":"The ancillary file nlmMultipoles.m is a useful addition, but the main text does not specify the exact format of the data or the conventions (e.g., the normalization factor N^{ℓm}_2 and the prefactor in Eq. (23)). A short README or a comment in the file would make the file self-explanatory.","section":"References"},{"comment":"The companion papers [82] and [87] are listed as 'to appear'; this is understandable but makes it impossible for the reader to check the deferred details at the time of submission. Please provide arXiv numbers when available.","section":"Introduction"}],"recommendation":"major_revision","confidential_remarks":"The paper is likely correct and of high interest, but the exact-in-velocity result is not verifiable from the preprint alone because the master-integral evaluation is deferred to a companion paper. This is a fixable issue—the authors can supply the missing details or an ancillary derivation file. I recommend major revision rather than acceptance at this stage, primarily for this verifiability reason. The claim about non-coherent states in the conclusions is also currently unsupported and should be tempered or justified."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Worth reading: this is the first all-orders-in-velocity result for the O(G^3) nonlinear memory multipoles in two-body scattering. The setup is standard for the amplitudes/PM program: the memory is traced to the low-frequency pole of the waveform, expressed through the Weinberg soft factor, and computed with reverse unitarity and IBP reduction. The final formula has a clean closed form with log and arccosh terms, and the low-velocity expansion is checked against an independent 9PN calculation [87] and against the leading small-v results [88,89]. Those checks are genuine and give me real confidence in the structure of the answer.\n\nThe main caveat is verifiability, not plausibility. The route from Eq. (15) to the polynomials in Table I and Appendix III relies on an IBP reduction and a canonical differential equation solution in the soft region, with details deferred to [82]. That reference is listed as “to appear,” and the diagrams in Eqs. (12)–(14) are not fully legible in this preprint, so the soft-limit input cannot be inspected directly. The small-v checks exercise the Taylor expansion around sigma=1, not the full finite-sigma functions; they would catch many but not all classes of error. I see no reason to suspect the coefficients are wrong — the method is mature and the authors have a strong track record — but the headline numbers are not independently checkable from this preprint alone.\n\nOne additional overreach: the conclusions claim the calculation shows that classical radiation is not described by a non-squeezed coherent state, contrary to previous expectations. That is a defensible inference only if the O(G^3) memory mode is connected to a specific coherence property of the quantum state. The paper does not show that connection; it shows the classical waveform has a nonlinear contribution. The discrepancy with [93,94] may be real, but the argument is sketched rather than demonstrated.\n\nOverall, the physics is important, the method is sound, and the cross-checks are meaningful. The paper deserves a serious referee. The referee should insist on [82] being available with the full IBP and master-integral details, or on the authors including enough in a revised version that the finite-sigma evaluation can be checked. The coherent-state remark should be toned down or substantiated. With those changes, the result should be accepted. I would cite it in my own work.","headline":"First exact-in-velocity computation of the O(G^3) nonlinear gravitational memory multipoles; the result is credible and well cross-checked, but the core master-integral evaluation is deferred to a companion paper.","tokens_in":17039,"tokens_out":2172,"would_cite":true,"duration_ms":24558,"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":"The paper computes the nonlinear gravitational memory waveform for two-body scattering exactly in velocity at the leading post-Minkowskian order.","keywords":["gravitational memory","nonlinear memory","post-Minkowskian expansion","scattering amplitudes","reverse unitarity","gravitational waveform","soft graviton theorem","two-body scattering"],"falsifier":"Compute the same $\\mathcal{O}(G^3)$ nonlinear memory multipoles by an independent classical method for a hyperbolic encounter at a nonzero velocity, for example by solving the relevant soft-graviton equations directly, at $\\sigma=2$, and compare with Table I; any mismatch in the polynomial coefficients $f_i^{\\ell m}$ would falsify the formula.","tokens_in":16002,"feed_emoji":"🌊","tokens_out":13630,"duration_ms":119928,"temperature":0.7,"pith_summary":"The paper aims to establish the first exact-in-velocity computation of nonlinear gravitational memory in the post-Minkowskian expansion: the permanent displacement left after two compact objects scatter, sourced by gravitons emitted by the gravitational waves themselves. Working at third order in Newton's constant, the leading order at which nonlinear memory appears, the paper derives closed-form expressions for the memory multipoles for arbitrary relative speed. The expressions are checked against post-Newtonian slow-motion expansions up to 9PN relative order and against earlier leading-velocity results. If correct, this completes the non-analytic-in-frequency part of the $\\mathcal{O}(G^3)$ multipolar waveform and supplies a benchmark for future waveform calculations in general relativity.","feed_headline":"Computing nonlinear gravitational memory exactly for the first time","feed_subtitle":"The first exact-in-velocity result for nonlinear memory, completing the low-frequency O(G^3) waveform.","key_machinery":"Multipoles $\\delta F^{\\ell m}$ of the nonlinear memory tensor, defined by projecting $\\delta F^{\\mu\\nu}$ onto spin-weighted spherical harmonics in the rest frame of one body, are the central objects. They are computed through the reverse-unitarity representation of $\\delta F$ as a three-particle cut of a two-loop integral, Eq. (15), with the observer direction encoded by extra powers of $(k\\cdot n)^{-1}$. The resulting integrals are reduced by integration-by-parts identities to a set of master integrals in the soft region, solved with canonical differential equations; the output is the velocity function $\\mathcal{F}^{\\ell m}(\\sigma)=(f_1^{\\ell m}+f_2^{\\ell m}\\log((\\sigma+1)/2)+f_3^{\\ell m}\\arccosh\\sigma/\\sqrt{\\sigma^2-1})/(\\sigma^2-1)^{\\ell/2}$, with $f_i^{\\ell m}$ rational polynomials in $\\sigma$.","core_discovery":"On its own terms, the paper's central discovery is that the nonlinear gravitational memory multipoles for two-body scattering are exactly computable at $\\mathcal{O}(G^3)$ in the post-Minkowskian expansion, for arbitrary relative velocity. The result takes the master form $\\delta F^{\\ell m}=G^3\\pi^2 m_1^2 m_2^2 b^{-3}(\\sigma^2-1)^{-3/2}\\, i^\\ell \\mathcal{N}_2^{\\ell m} \\, \\mathcal{F}^{\\ell m}(\\sigma)+O(G^4)$, with $\\mathcal{F}^{\\ell m}(\\sigma)$ built from polynomials in $\\sigma$ plus $\\log((\\sigma+1)/2)$ and $\\arccosh\\sigma/\\sqrt{\\sigma^2-1}$. The paper extracts $\\delta F^{\\ell m}$ from the soft-graviton limit of the two-loop amplitude using a reverse-unitarity cut representation, evaluates the master integrals by canonical differential equations, and tabulates the polynomial coefficients for $\\ell\\le5$. It reports that all odd $\\ell+m$ multipoles vanish and that the small-velocity expansion reproduces the known 9PN and leading-velocity results. It further concludes that this soft-limit origin implies classical gravitational radiation is not fully captured by a non-squeezed coherent state, with possible classical correlations between detectors.","pith_inferences":["If correct, the exact-in-velocity polynomials can be expanded to arbitrary post-Newtonian order, yielding higher-order small-velocity data beyond the 9PN check reported in the paper.","If correct, the coherent-state remark can be turned into a quantitative prediction: a two-detector correlation function built from these multipoles would let detectors test the classical, nonlinear origin of the effect.","If correct, the same reverse-unitarity and master-integral machinery should apply to multipoles with spins, since the derivation does not appear to use spinlessness."],"forward_implications":["The nonlinear memory multipoles at $\\mathcal{O}(G^3)$ are now known as exact functions of velocity, not just as slow-motion expansions.","The gauge-invariant non-analytic-in-frequency part of the $\\mathcal{O}(G^3)$ multipolar waveform is complete, so future amplitude or classical calculations of the waveform can be checked against these formulas.","The small-velocity expansion of the new formula reproduces the independent 9PN results and earlier leading-velocity results, providing a cross-check of both approaches.","Because nonlinear memory appears at $\\mathcal{O}(G^3 v)$, it is suppressed relative to the leading quadrupole by 2.5 post-Newtonian orders in the scattering setup.","The soft-limit origin of the result implies classical radiation is not described by a non-squeezed coherent state, suggesting classical correlations between gravitational-wave detectors."],"supporting_citations":[{"why":"This work supplies the amplitude-based representation of the gravitational waveform that turns the two-loop computation into the memory integral.","marker":"[39]"},{"why":"This work establishes the leading soft-graviton theorem that identifies the $1/\\omega$ pole with memory and includes graviton emission from gravitons, which is exactly the nonlinear part computed here.","marker":"[67]"},{"why":"This work provides the reverse-unitarity and balance-law framework used to write the nonlinear memory as a three-particle cut and to connect it to radiated energy-momentum.","marker":"[14]"},{"why":"This work supplies the two-loop master-integral families in the elastic case on which the present soft-region reduction is built.","marker":"[78]"},{"why":"This work provides the canonical differential-equation method and soft-region master integrals used to evaluate the cut integrals exactly in velocity.","marker":"[16]"},{"why":"This work introduces the $S=e^{iN}$ N-matrix elements used to organize the two-loop waveform into soft-graviton building blocks.","marker":"[75]"}],"fun_headline_variants":["Exact nonlinear memory from two-body scattering at leading PM","Nonlinear memory waveform computed exactly for the first time","Soft gravitons give exact nonlinear memory at O(G^3)","First exact-in-velocity results for nonlinear gravitational memory","Nonlinear memory at leading order in post-Minkowskian expansion"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the two-loop amplitude's leading low-frequency terms, combined with the reverse-unitarity cut representation, capture the full classical nonlinear memory at this order; the paper defers the detailed integral reduction and boundary conditions to a companion paper (after Eq. (23)), so the tabulated coefficients cannot be independently checked from the preprint alone.","fun_headline_variants_meta":{"raw":{"variants":["Exact nonlinear memory from two-body scattering at leading PM","Nonlinear memory waveform computed exactly for the first time","Soft gravitons give exact nonlinear memory at O(G^3)","First exact-in-velocity results for nonlinear gravitational memory","Nonlinear memory at leading order in post-Minkowskian expansion"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000501,"raw_usage":{"total_tokens":2454,"prompt_tokens":954,"completion_tokens":1500,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":570,"completion_tokens_details":{"reasoning_tokens":1416}},"tokens_in":570,"tokens_out":1500,"duration_ms":12257,"temperature":1.0,"reasoning_tokens":1416,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T22:43:24.069999+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the same $\\mathcal{O}(G^3)$ nonlinear memory multipoles by an independent classical method for a hyperbolic encounter at a nonzero velocity, for example by solving the relevant soft-graviton equations directly, at $\\sigma=2$, and compare with Table I; any mismatch in the polynomial coefficients $f_i^{\\ell m}$ would falsify the formula.","supporting_citations":[],"review_version":1}