{"id":"ef37025f-c274-47db-b0b4-17dfbb20181f","arxiv_id":"2607.27133","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Classical ULDM condensate decouples from GW propagation; squeezing-induced parametric resonance of primordial tensor modes is ≲10^{-12} for non-relativistic ULDM at equality.","lead":"A first-principles quantum field theory calculation shows that a classical ultralight dark-matter condensate does not affect gravitational-wave propagation, contrary to earlier claims. Quantum squeezing can drive a parametric resonance, but the amplification is negligible under realistic conditions.","discovery_kind":"first_principles","skeptic_critique":{"model":"moonshotai/kimi-k3","headline":"The condensate-decoupling result is structurally sound, but the ≲10^{-12} resonance bound inherits all of its control from the mid-point/first-order phase expansion and the stricter cutoff condition (5.4); outside that regime the Mathieu analysis is not shown to be complete.","rationale":"I read the paper in good faith and stress-tested both pillars of the strongest claim. (1) Condensate decoupling: the cancellation of the condensate 4-point self-energy against gravitational-sector terms is driven by background symmetries (eq. 2.53), and I checked that every tensor structure in the non-local condensate self-energies (3.43)-(3.44) carries either η_{ρσ} or a δ^0 on the (ρσ) side, so all vanish under TT contraction — consistent with their absence from the TT equation (4.8). This claim does not depend on the adiabatic or mid-point approximations at leading order and I find no soft spot. (2) The 10^{-12} bound: I independently re-derived the Floquet-crossing exponent and recovered eqs. (5.49) and (5.52) including the non-obvious k^{17/2}, m^{11/2} scalings, which increases confidence in the analytic chain. The remaining control parameter is exactly the reader's: the first-order phase expansion for I_2^S (condition 5.4) plus the mid-point treatment of non-local terms (3.12, 5.13; App. C-D). The paper is transparent about this and even gives the exact unexpanded result (C.16), whose dephasing factor shows precisely how the approximation fails when (5.4) is violated; the authors also note higher-order terms relax (5.4) toward (5.2). So the concern limits the parameter region over which the quantitative bound is demonstrated, but does not impeach the result within the stated regime, and the bound itself is conservative (pressure saturating w~1 at equality). This matches the reader's CONDITIONAL/HIGH-confidence assessment, and I see no basis to move the verdict in either direction: the structural claim is stronger than CONDITIONAL suggests, but the headline number is only as controlled as (5.4), so CONDITIONAL remains right. The proposed numerical integration of (4.10) with the exact phase would settle the issue decisively and doubles as the paper's own suggested future check.","tokens_in":89376,"tokens_out":8869,"duration_ms":172918,"concrete_test":"Numerically integrate the full graviton equation (4.10) — including the non-local retarded self-energy from eqs. (3.35)-(3.40) evaluated with the exact hypergeometric adiabatic phase (3.11), i.e. without the mid-point assumption (3.12) — for a benchmark pure squeezed state with m = 10^{-22} eV and a power-law spectrum whose cutoff saturates condition (5.4), k_UV/a_eq = √(m H_eq). Track the maximally resonant mode k ≈ m a_eq across matter domination and compare the measured log-enhancement to the analytic estimate (5.49)/(5.52). If the numerical enhancement exceeds the local-Mathieu prediction by more than the estimated k²_UV/(a²mH) correction, the suppression claim and the 10^{-12} bound fail at the boundary of their validity regime; agreement would validate the mid-point approximation. A cheaper analytic precursor: recompute I_2^S keeping the n=2 phase term in the expansion (C.9) and re","verdict_should_be":"UNCHANGED","load_bearing_attack":"The two halves of the strongest claim have very different footing. The condensate decoupling is robust: the 4-point condensate self-energy cancels gravitational-sector terms via the background Noether-Ward identity (eqs. 2.52-2.53), and the non-local condensate 3-point structures (3.43-3.44) all carry η_{ρσ} or δ^0 indices that vanish when contracted with a transverse-traceless h_{ρσ} — which is why no condensate term survives in eq. (4.8). This is algebra, not approximation, and the tension with refs. [15-17] is a scope difference (homogeneous condensate vs. halos/scalar-tensor), not an inconsistency. I also verified the narrow-resonance crossing integral independently: linearizing the detuning drift ∆ ≈ −2m_S²ℋ(η−η_res) and integrating µ_+ = (1/2m_S)√(g_1²−∆²) reproduces the exponent π g_1²/(8 ω_res³ℋ) of eq. (5.49), and the substitution chain through a_res = 3k/m and H ∝ a^{-3/2} reproduces the m^{11/2} z_eq^{3/2}/(k^{17/2} H_eq) scaling of eq. (5.52). The final 10^{-12} figure is a generous upper bound (P_DM set by w_DM ≲ 1 at equality). The genuinely load-bearing point is the one the reader flagged: the local squeezing mass I_2^S is evaluated with the amplitude at zeroth order and the phase at first order in k²/(a²m²) (App. C, eq. C.11-C.18), which requires k_UV/a_eq ≪ √(mH_eq) (eq. 5.4) rather than the natural non-relativistic condition k_UV/a ≪ m (eq. 5.2), and the neglect of the non-local self-energy rests on the same condition via the mid-point phase assumption (eq. 3.12) and the estimate (5.13). If (5.4) is only marginally satisfied, the dephasing factor [1+9k⁴_UV(t/a²−t_eq/a²_eq)²/m²]^{-(n_S+2j+1)/4} in the exact result (C.16) is no longer ≈1, the collective oscillation frequency m_S and its amplitude are modified, and the non-local 'relativistic correction' classification (5.13) loses its small parameter — so both legs of the quantitative bound sit on one approximation whose error is asserted, not computed, to be small.","agreement_with_reader":"agree"},"referee_report":{"model":"moonshotai/kimi-k3","summary":"The manuscript develops a first-principles quantum-field-theoretic treatment of linear cosmological perturbations during matter domination sourced by ultralight scalar dark matter in a general Gaussian state (condensate, occupation number, and squeezing). Working in the Schwinger–Keldysh/2PI formalism with a classically treated graviton, the authors derive a closed linearized graviton equation sourced by background matter correlators, renormalize the one-loop vacuum self-energy in dimensional regularization with the counterterm action (2.45), and evaluate state-dependent contributions in the adiabatic (WKB) regime using a mid-point prescription for non-local terms. Two main results are claimed: (i) after gauge fixing, the homogeneous ULDM condensate has no effect on gravitational-wave propagation, because its four-point self-energy cancels against gravitational-sector terms rewritten via the background Einstein equation (eqs. 2.50–2.53, 4.8) — a consequence of background FLRW symmetries; (ii) the time-dependent effective graviton mass induced by mode squeezing produces a Mathieu-type parametric resonance for specific tensor modes, but the relative enhancement is bounded at ≲10⁻¹² for non-relativistic ULDM at matter–radiation equality with power-law spectra in the range m∼10⁻²¹–10⁻²⁴ eV (eq. 5.55). Scalar-potential equations are derived but their analysis is deferred.","tokens_in":89899,"tokens_out":7350,"duration_ms":368164,"significance":"If the results hold, the paper makes two contributions of note. First, it proves that a homogeneous ULDM condensate drops out of the linearized GW equation — not as an approximation but as a structural cancellation between the condensate four-point self-energy and gravitational-sector terms enforced by the background symmetries (a Noether–Ward identity). This corrects, or at least sharply delimits, earlier claims of condensate-driven GW resonance, and clarifies that any ULDM imprint on tensor modes must come from two-point statistics. Second, it identifies a genuinely quantum mechanism — squeezing-induced oscillatory graviton mass driving parametric resonance — and delivers a falsifiable, and as it turns out strongly negative, quantitative bound (≲10⁻¹²) on the enhancement under stated assumptions. The work ships an explicit dimensional-regularization renormalization of the graviton self-energy on FLRW with a full counterterm action, causal (retarded-only) in-in dynamics, and a resonance exponent whose derivation can be checked independently (the narrow-resonance crossing integral and the m^{11/2} z_eq^{3/2} k^{−17/2} scaling of eq. (5.52) reproduce correctly). The negative result,","major_comments":[{"comment":"§5, eq. (5.4) vs (5.2); App. C, eqs. (C.9)–(C.11): the evaluation of I^S_2 retains the phase only to first order in k²/(a²m²), requiring k_UV/a_eq ≪ √(mH_eq) — for m ~ 10⁻²² eV roughly three orders of magnitude tighter than the natural non-relativistic condition (5.2). The relaxation of (5.4) to (5.2) via higher-order phase terms is asserted but not quantified: higher orders add k-dependent frequency dispersion that can dephase the resonance and change the Mathieu analysis. The abstract and §6 state the 10⁻¹² bound for 'non-relativistic ULDM at equality' without this qualification. Please (a) state the regime of validity of (5.55) explicitly in the abstract/conclusions, and (b) quantify the error or dephasing when (5.2) holds but (5.4) fails.","section":"§5, eq. (5.4); App. C"},{"comment":"§5, eqs. (5.7)–(5.13); App. D: the neglect of the non-local self-energy — needed to reduce (4.10) to the Mathieu equation (5.42) — rests on the adiabatic ansatz (5.7) with an undetermined exponent α, on ω(k;t̄)≈ω(k;t), and on ∂_t f ~ H f. Near the special frequencies (5.11)–(5.12) the suppression is only k²_UV(t−t_eq)/(a²m), whose time domain of validity is not stated, and the possible overlap of these non-local resonant frequencies with the local Mathieu band (k≈am/3) is not discussed. Since this is the step that converts the integro-differential equation into the solvable local one, please give explicit bounds on the time window and α-independence of the estimate, and state the resulting uncertainty on (5.55).","section":"§5, eqs. (5.7)–(5.13); App. D"},{"comment":"§5.2.3, eqs. (5.54)–(5.55): the headline number 10⁻¹² is obtained by setting P_DM,0 ~ 10⁻¹⁸ eV⁴, i.e. w ~ 1 at equality. But w ~ 1 at equality violates the non-relativistic assumption (5.21) under which the power-law expansions of I^N_2 and I^S_2 (eqs. (5.3)–(5.5)) — and hence the Mathieu equation itself — are derived. The bound is presented as 'generous', but strictly it is an extrapolation outside the controlled regime; a self-consistent evaluation with w_eq ≪ 1 would yield a parametrically smaller enhancement. Please discuss this self-consistency and, if possible, quote the bound obtained at the edge of the controlled regime (e.g. w_eq ≲ 0.1).","section":"§5.2.3, eqs. (5.54)–(5.55)"}],"minor_comments":[{"comment":"m_S is defined twice with different content: eq. (5.10) (coefficient n_S+7, arising from the p⁶-weighted non-local integral σ_S in (D.23)–(D.24)) and eq. (5.41) (conformal-time definition entering the Mathieu equation, tied to the n_S+5 local frequency of (5.5)). Please distinguish the two symbols.","section":"§5, eqs. (5.10) and (5.41)"},{"comment":"The abstract and §1 state the condensate-decoupling result is 'contrary to previous claims' [15–17]. Those works treat inhomogeneous halo configurations and scalar-tensor couplings, whereas the cancellation (2.53)/(4.8) holds for a spatially homogeneous condensate on FLRW. An explicit sentence delineating the scope would prevent the conflict from being overstated.","section":"Abstract; §4.1"},{"comment":"The choice c^f_4 = −4F_m/(16π)² (below (3.22) and (B.40)) nullifies the vacuum energy-momentum tensor (3.46)–(3.47). This is a renormalization (cosmological-constant scheme) condition and should be identified as such, since it is a physical input rather than a derived result.","section":"§3.3.1–3.4, eqs. (3.46)–(3.47)"},{"comment":"Typos/notation: 'FLR W' (spurious space) throughout; 'Lichnerowitz' (App. A, above (A.4)) vs 'Lichnerowicz'; 'on a on a perturbed' (§1); 'the vertices also acquires this index' (App. A); 'Using the the definition' (§5.2.2); '4X_{t=i}' in (3.39); 'palindrome function' in App. B.1.4 presumably means 'multivalued function'; Fig. 1 caption 'one graviton leg needs to be truncated' is unclear.","section":"Various"},{"comment":"Eq. (2.53): it would help to state explicitly that the cancellation is an on-shell identity at background level, i.e. it uses the background Einstein equation (2.32); this is the precise sense in which it is a Noether–Ward consequence, as invoked again in §4.1.","section":"§2.4, eq. (2.53)"},{"comment":"Eq. (5.53): the numerical range k ≳ ma_eq ∼ 10⁻²⁴–10⁻²⁷ eV and the mapping to λ ≲ 10²–10⁵ Mpc should be double-checked for units and for consistency with the sub-horizon requirement kη≫1 used in the Floquet treatment; a short table of the resonant k-window versus m would aid the reader.","section":"§5.2.3, eq. (5.53)"},{"comment":"Given the relevance to PTA constraints, the discussion of [65, 69, 70] in §6 could include one or two more recent PTA–ULDM analyses so readers can locate the scalar-sector claim (left to future work) in the current literature.","section":"§6"}],"recommendation":"minor_revision","confidential_remarks":"The manuscript is very long (95 pp., four heavy appendices). I verified the structural cancellation (2.52)–(2.53), the TT projection in (4.8), and the narrow-resonance exponent chain (5.47)–(5.52) by independent recomputation; these check out. I did not verify the full tensor algebra of (3.35)–(3.40) or App. B line by line; editors may wish to weigh that the load-bearing results rest on the parts I could check. The \"contrary to previous claims\" framing toward [15–17] is a scope difference, not an error, but the authors should be asked to say so explicitly. No concerns about citation practice or scope fit for JHEP."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"The load-bearing result is clean: after gauge-fixing, the homogeneous ULDM condensate cancels from the linear graviton equation. The 4-point condensate self-energy is killed by the background FLRW terms once you rewrite the Lichnerowicz operator with ρ and P (2.52–2.53), and the remaining 3-point pieces carry indices that vanish on a TT tensor. That is algebra plus the Noether–Ward identity, not an approximation, and it directly contradicts the classical-condensate claims in [15–17] for this setting (homogeneous background, not halos). They also give a controlled Floquet estimate that squeezing-driven Mathieu resonance stays ≲10^{-12} for standard fuzzy-DM masses once you impose non-relativistic conditions at equality.\n\nWhat is new is the full 2PI in-in setup for a generic Gaussian state (condensate + N_k + S_k), the explicit adiabatic self-energy (local + non-local, renormalized with the counterterm action 2.45), and the reduction to a Mathieu equation whose narrow-resonance crossing they integrate analytically. The vacuum renormalization is done carefully; the occupation/squeezing integrals are written so you can re-derive them. Circularity is low: Ω_DM and w_DM ≪ 1 fix N_0, n_N; S_0, n_S are free but bounded by purity.\n\nSoft spot, in proportion: the quantitative bound inherits its control from the mid-point phase assumption and the stricter cutoff k_UV/a_eq ≪ √(m H_eq) (5.4) needed to keep the first-order phase error small and to classify non-local terms as O(k_UV²/a²m²) corrections (App. C–D, 5.13). The natural non-relativistic condition is only k_UV/a ≪ m. If (5.4) is marginal the dephasing factor in the exact I_2^S and the non-local estimate both move, so the Mathieu analysis is not fully closed. They flag this and leave a numerical check of the non-locals for later; that is honest, not fatal. Scalar potentials are derived but not analyzed.\n\nThis is for people working on quantum corrections to cosmological GWs or on whether ULDM quantum pressure can source PTA/CMB signals. The condensate cancellation alone is worth having on the record. I would send it to referees; the formal core is solid enough to deserve the scrutiny, and the bound is useful even with the stated caveat. Engage if you care about the ULDM–GW interface; skip if you only want classical halo phenomenology.","headline":"Solid 2PI derivation showing the homogeneous ULDM condensate drops out of linear GW propagation; the ≲10^{-12} squeezing-resonance bound is real but sits on a stricter mid-point/cutoff assumption than the natural non-relativistic one.","tokens_in":88423,"tokens_out":679,"would_cite":true,"duration_ms":16185,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.5","headline":"Classical ultralight dark-matter condensates do not alter gravitational-wave propagation; squeezed-state quantum pressure can, but the resonant boost is tiny.","keywords":["ultralight dark matter","cosmological perturbations","gravitational waves","parametric resonance","squeezed states","2PI effective action","adiabatic approximation","matter-dominated era"],"falsifier":"A numerical evaluation of the full non-local retarded self-energy (without the mid-point approximation) for a pure squeezed power-law spectrum that yields a resonant enhancement larger than ∼10^{-12} for any mode that enters the horizon after equality in the stated mass window.","tokens_in":87835,"feed_emoji":"🌊","tokens_out":1008,"duration_ms":22840,"temperature":0.7,"pith_summary":"During the matter-dominated era, dark matter drives the growth of primordial perturbations. Ultralight scalar dark matter is usually treated as a classical condensate. This paper builds a first-principles quantum field theory for how a general Gaussian state of that field sources linear cosmological perturbations, keeping the quantum nature of the matter. After gauge fixing, the classical condensate drops out of the gravitational-wave equation entirely, contrary to earlier claims. What remains is a time-dependent effective mass for the graviton that comes from the quantum pressure of a squeezed state. That mass can produce parametric resonance in specific primordial gravitational-wave modes. For non-relativistic ultralight dark matter at matter-radiation equality and a power-law squeezing spectrum, the relative amplification stays below about 10^{-12} across the mass window 10^{-21} to 10^{-24} eV, too small for foreseeable detection.","feed_headline":"Classical ULDM condensates leave gravitational waves untouched","feed_subtitle":"Squeezed-state quantum pressure can resonate, but the boost stays below 10^{-12}","key_machinery":"The closed semiclassical graviton equation obtained from the 2PI effective action on the Schwinger–Keldysh contour, reduced in the adiabatic (WKB) regime with the mid-point prescription for non-local phases. After longitudinal-gauge fixing the transverse-traceless sector becomes a Mathieu equation whose oscillatory mass is set by the local squeezing integral I^S_2.","core_discovery":"After gauge fixing, the homogeneous classical ULDM condensate has no influence on gravitational-wave propagation: its four-point self-energy cancels against corresponding local terms in the Einstein tensor by background symmetries. The residual effect is a time-dependent graviton effective mass generated by the quantum pressure of the squeezed state, which can drive parametric resonance. Under a power-law squeezing spectrum and non-relativistic conditions at equality, that resonant growth is negligible (relative enhancement ≲ 10^{-12}) for masses m ∼ 10^{-21}–10^{-24} eV.","pith_inferences":["If the mid-point approximation is only marginally valid near equality, residual non-local contributions could accumulate over many Hubble times and slightly loosen the 10^{-12} bound without overturning the qualitative cancellation of the condensate.","The same 2PI-plus-adiabatic pipeline can be ported to the radiation-to-matter transition or to super-horizon modes, where the condensate cancellation may no longer hold and larger effects could appear.","A pure squeezed initial state is the most optimistic case for detection; any realistic decoherence during radiation domination would push the signal still lower, reinforcing the null result for tensors."],"forward_implications":["A purely coherent (classical-condensate) treatment of ULDM cannot produce an observable imprint on primordial gravitational waves at linear order.","Any detectable quantum signature in tensor modes during matter domination would have to come from the two-point statistics (occupation number or squeezing), not from the one-point condensate.","The pure squeezed state supplies an absolute upper bound on resonant growth; mixed states produce even weaker resonances.","Scalar Bardeen potentials remain coupled to the condensate and to squeezing, so they may still carry distinctive oscillatory signatures even when tensors do not.","Existing PTA-style bounds that rely on a classical oscillating condensate sourcing gravitational waves need re-examination in light of the cancellation."],"fun_headline_variants":["Classical ULDM condensate leaves GW propagation untouched","Squeezed ULDM quantum pressure drives negligible GW resonance","ULDM classical condensate cancels from graviton dynamics","Gauge fixing erases classical ULDM effect on gravitational waves","Parametric resonance from ULDM quantum pressure stays below 1e-12"],"cache_read_input_tokens":128,"weakest_assumption_plain":"The mid-point approximation that replaces non-local adiabatic phases by their values at the average time, together with the stricter ultraviolet cutoff needed to keep the first-order phase error small; if that approximation fails, neither the claimed suppression of non-local terms nor the Mathieu resonance analysis is controlled.","fun_headline_variants_meta":{"raw":{"variants":["Classical ULDM condensate leaves GW propagation untouched","Squeezed ULDM quantum pressure drives negligible GW resonance","ULDM classical condensate cancels from graviton dynamics","Gauge fixing erases classical ULDM effect on gravitational waves","Parametric resonance from ULDM quantum pressure stays below 1e-12"]},"model":"grok-4.5","effort":"low","cost_usd":0.004359,"raw_usage":{"total_tokens":1339,"prompt_tokens":810,"num_sources_used":0,"completion_tokens":65,"cost_in_usd_ticks":43588000,"prompt_tokens_details":{"text_tokens":810,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":464,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":810,"tokens_out":65,"duration_ms":8092,"temperature":1.0,"reasoning_tokens":464,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-30T11:47:54.271624+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"A numerical evaluation of the full non-local retarded self-energy (without the mid-point approximation) for a pure squeezed power-law spectrum that yields a resonant enhancement larger than ∼10^{-12} for any mode that enters the horizon after equality in the stated mass window.","supporting_citations":[],"review_version":1}