{"id":"bac1492c-e2c2-44c6-8767-9de6804985c3","arxiv_id":"2512.21108","paper_version":2,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":2,"one_line_summary":"Axion-photon conversion in random Gaussian magnetic fields fixes the expectation values and variances of photon Stokes parameters, including a circular-polarization signal from helical fields.","lead":"Photons traveling across the universe can convert into hypothetical axion particles when they pass through magnetic fields. This paper works out how that conversion changes light's brightness and polarization when the magnetic fields are random, giving new ways to search for axions.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Absolute value in Eq (3.26) flips the sign of β and thus Exp[V] when m_a<m_pl","rationale":"The paper is a careful and mostly sound derivation, and the reader's identification of Gaussianity as the key assumption is reasonable but not the most load-bearing issue. The Gaussian assumption is stated explicitly and is standard for stochastic magnetic-field models; within that assumption the Wick contractions in Appendix A are correct, and the consistency relations are algebraic consequences of the variance formulas. However, Eq (3.26) introduces an absolute value in the phase difference Π that does not follow from the equations of motion and is not a harmless convention: for m_a < m_pl and low frequencies, the physical detuning δ is negative, and the absolute value changes the sign of the helical term β. Since the central claim explicitly includes Exp[V] = -β, this is an internal inconsistency for part of the parameter space rather than just a limitation of an external assumption. It does not affect the variance formulas or the consistency relations (which involve only squares of β), nor does it affect the intensity or linear-polarization variances; the issue is confined to the sign of the predicted circular polarization. The fix is straightforward: define Π as the signed difference or state the assumption m_a > m_pl. Because the paper's main new qualitative claim—the circular-polarization peak—is affected in a significant regime, I would adjust the verdict from unconditional ACCEPT to CONDITIONAL acceptance pending this correction or an explicit domain restriction.","tokens_in":31027,"tokens_out":30312,"duration_ms":293941,"concrete_test":"Re-derive the Born amplitudes in Eq (3.25) using the signed phase difference δ(ω) = (m_a^2 - m_pl^2)/(2ω) + χ_CMB ω instead of the absolute value in Eq (3.26). Then compute β = (g_aγγ/2)^2 ∫ d^3k/(2π)^3 I(k_z;δ) k̂_z P_aB(k) for a concrete case with m_a < m_pl and ω below the crossing where δ < 0, and compare the sign of Exp[V] = -β with that obtained from Eq (4.4). If the sign flips, the absolute value in Eq (3.26) is the source and the central claim requires either the restriction m_a > m_pl or a sign-corrected definition of β.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The paper's central claim includes Exp[V] = -β, with β defined in Eq (4.5) through the kernel I(k_z;ω). But Eq (3.26) defines Π(ω) = |m_a^2 - m_pl^2|/(2ω) + χ_CMB ω, whereas the actual phase difference obtained from Eqs (3.13)-(3.14) is δ(ω) = (m_a^2 - m_pl^2)/(2ω) + χ_CMB ω. For m_a < m_pl and sufficiently small ω, δ is negative, and the absolute value replaces δ with |δ| = -δ. Because I(k_z; -|δ|) = I(-k_z; |δ|), the angular integral defining β changes sign under k → -k: β(-|δ|) = -β(|δ|). Hence the predicted expectation value of the circular polarization has the wrong sign on this branch. The variances and consistency relations depend only on α^2, β^2, γ^2, so they are unaffected, but the paper's claim of a peak in Exp[V] with a definite sign is not correct as stated for m_a < m_pl. The formulas are valid as written only if m_a > m_pl, or if β is redefined with the physical signed Π.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript develops a statistical treatment of axion-photon conversion in stochastic magnetic fields. It considers a photon propagating through a statistically homogeneous, isotropic Gaussian magnetic field, treats the axion-photon mixing in the Born approximation, and derives the ensemble mean and variance of the Stokes parameters after propagation. For initially unpolarized photons, the mean conversion probability and mean circular polarization are expressed in terms of two spectral integrals α and β, while the variances are quadratic forms in α, β, and γ. The paper also derives spectrum-independent consistency relations among these statistical moments and illustrates the frequency dependence with broken power-law spectra, finding a peak in the mean circular polarization for helical fields.","tokens_in":31358,"tokens_out":9087,"duration_ms":88632,"significance":"The strength of the paper is its analytic control: the Wick-contraction calculation in Appendix A is explicit, the convolution kernels are given in closed form in Appendix B, and the consistency relations are parameter-free consequences of the Gaussian and Born assumptions. If correct, the formalism provides a useful tool for estimating polarization signatures of axion-photon conversion in cosmic magnetic fields and gives falsifiable relations that can be checked in simulations or observations. The helical-field extension goes beyond existing domain-like and statistical treatments. However, a sign error in the definition of the oscillation frequency affects the predicted circular polarization on part of the parameter space, which is one of the paper's headline claims.","major_comments":[{"comment":"The absolute value in Eq. (3.26) is not equivalent to the physical phase difference. From Eqs. (3.13)–(3.14), Π_a−Π_γ = (m_a^2−m_pl^2)/(2ω)+χ_CMBω, which is negative for m_a<m_pl and small ω. Since I(k_z;ω) in Eq. (4.6) is even in Π−k_z, replacing Π by −Π is equivalent to k_z→−k_z, and the angular integral defining β in Eq. (4.5) changes sign. Hence Exp[V]=−β flips sign on this branch. The variances and consistency relations involve only α², β², γ² and are unaffected. The numerical figures in Sec. V use only |m_a^2−m_pl^2| and do not specify the sign; the statement in Sec. V C that V is opposite to the magnetic helicity is therefore not generally valid. Please use the signed Π throughout and qualify the circular-polarization predictions by the sign of m_a^2−m_pl^2, or explicitly restrict to m_a>m_pl.","section":"Eq. (3.26), Secs. III B, IV B, V C"},{"comment":"The numerical illustrations use representative parameters B_*∼nG, d∼100 Mpc, g_aγγ∼10^−12 GeV^−1, for which the first-order Born amplitude gBd is of order 0.1 and the truncation of the perturbation series is not obviously justified. Since the analytic results are explicitly leading order, the paper should either state the weak-coupling regime in which they apply or quantify the expected size of the next-order correction. This is not a problem for the formal derivation, but it is relevant for the observational conclusions drawn from the figures.","section":"Sec. V"}],"minor_comments":[{"comment":"The phrase \"statically unpolarized\" should be \"statistically unpolarized\" (or \"initially unpolarized\").","section":"Sec. VI"},{"comment":"The claim that the circular polarization is opposite to the magnetic helicity needs a qualifier specifying the sign of m_a^2−m_pl^2; see Major Comment 1.","section":"Abstract and Sec. V C"},{"comment":"The δ² terms are displayed and then immediately set to zero; consider removing the intermediate equalities to avoid confusing the reader about a potentially nonvanishing quantity.","section":"Eq. (4.4)"},{"comment":"The definition of ω_eq uses |m_a^2−m_pl^2|; the caption should state this explicitly so that the folding of the frequency axis is not misinterpreted as a physical symmetry.","section":"Fig. 2 caption"},{"comment":"The consistency relations are clearly derived; it would be helpful to state once more that they hold only under Gaussian statistics and the Born approximation, as these are the assumptions that enter through Eq. (4.4).","section":"Sec. IV C"}],"recommendation":"major_revision","confidential_remarks":"The paper is a solid analytic contribution and the derivation is reproducible from the manuscript. The sign error in Π(ω) is localized and fixable, so I do not recommend rejection. However, because the sign of the circular polarization is one of the paper's headline predictions, the revision should be substantive rather than cosmetic. I would be happy to re-review after the sign issue is addressed."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First, the stress-test note is right. Eq (3.26) defines Π as |m_a^2 - m_pl^2|/(2ω)+χω, but Eqs (3.13)-(3.14) give Π_a - Π_γ = (m_a^2 - m_pl^2)/(2ω)+χω. Those differ when m_a < m_pl. The absolute value is not a harmless convention: the kernel I satisfies I(k_z;-Π)=I(-k_z;Π), so after the isotropic integral for β, β(-Π) = -β(Π). Since Exp[V] = -β, the predicted circular polarization flips sign on the low-frequency branch ω < ω_eq relative to the magnetic helicity. That is exactly the observable the abstract advertises. The variances and consistency relations are built from squares and are unaffected; the intensity is even in Π. So the core analytic results stand, but the 'peak structure with a definite sign' claim needs correction for m_a < m_pl.\n\nWhat is genuinely new is the full set of expectation values and variances of Stokes parameters for initially unpolarized photons in a homogeneous, isotropic Gaussian magnetic field, including the helical component. The Appendix A Wick-contraction calculation is explicit, the convolution kernels are given in closed form, and the three consistency relations are parameter-free and nontrivial. This is a real extension of the earlier intensity-only, non-helical treatments, and the connection to the authors' graviton-photon work is legitimate analogy, not a hidden input. The numerics are honest illustrative model work.\n\nMinor caveats: the Born approximation validity is not quantified for the Sec V parameter choices, and the Gaussian assumption is standard but stated. Neither bothers me much.\n\nBottom line: careful, useful paper with one load-bearing sign error in the headline circular-polarization result. The fix is small — use the signed mass difference or explicitly restrict to m_a > m_pl and give the sign-flip for the other case — and the variance formalism and consistency relations are worth publishing. I would send it to a referee, not desk-reject.","headline":"Useful new formalism for polarization statistics in axion-photon conversion, but a sign error in the phase Π flips the headline circular-polarization prediction for m_a < m_pl; the fix is small and the core results survive.","tokens_in":31838,"tokens_out":8367,"would_cite":true,"duration_ms":77589,"reading_group":"yes","serious_thinker":"no","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"For initially unpolarized photons crossing a random magnetic field, axion-photon conversion makes all Stokes-parameter statistics expressible through two magnetic-field power spectra, including new consistency relations that hold for any sp","keywords":["axion-photon conversion","stochastic magnetic fields","Stokes parameters","circular polarization","helical magnetic fields","magnetic-field power spectra","Born approximation","gamma-ray polarization"],"falsifier":"Take a set of many observed photon sources whose lines of sight traverse the same statistical magnetic-field region, measure their Stokes parameters, and compute the ensemble variance of Q and U; if the variances are unequal, or if the combination Var[1−I] − Var[Q] − Exp[V]² is measurably nonzero, the Gaussian-Born prediction is ruled out. Alternatively, a direct measurement of the four-point magnetic-field correlator would reveal non-Gaussianity and invalidate the Wick factorization at the heart of the derivation.","tokens_in":30910,"feed_emoji":"🌀","tokens_out":2647,"duration_ms":34030,"temperature":0.7,"pith_summary":"This paper tries to show that when initially unpolarized photons travel through a stochastic magnetic field, axion-photon conversion generically produces nonzero polarization variances — and, if the field is helically structured, a nonzero expectation value of circular polarization. The authors derive closed-form formulas for the expectation values and variances of the Stokes parameters I, Q, U, V in terms of the magnetic field's symmetric and antisymmetric (helical) power spectra, assuming a Gaussian, statistically homogeneous and isotropic field and a weak-coupling Born approximation. They then extract three consistency relations among these statistical quantities that do not depend on the specific shape of the power spectra, making them robust observational tests. A sympathetic reader would care because this turns high-energy photon observations of cosmological sources into a quantitative probe of axion parameters and magnetic-field helicity.","feed_headline":"Random fields polarize light via axion conversion","feed_subtitle":"Consistency relations tie photon polarization variances to magnetic-field spectra — a testable axion probe.","key_machinery":"The central object is the statistically homogeneous, isotropic Gaussian magnetic field, characterized by its two-point correlator in terms of a symmetric spectrum P_B(k) and an antisymmetric helical spectrum P_aB(k). The argument combines the Schrödinger-like mixing equation of axion-photon propagation with a first-order Born approximation, so the photon fields after propagation are linear in the magnetic field and the Stokes parameters become quadratic products. The ensemble statistics then reduce, via Wick-contracted four-point functions, to products of these two spectra weighted by two convolution kernels I(k_z;ω) and J(k_z;ω) — sinc-like integrals over the propagation path. The four inte","core_discovery":"For unpolarized photons in a Gaussian stochastic magnetic field, the ensemble mean and variance of every Stokes parameter are fully determined by the power spectra P_B(k) and P_aB(k) through the three integrals α, β, and γ (with a fourth integral δ vanishing identically). Explicitly, the expectation of fractional intensity loss is α, the expectation of circular polarization is −β, the linear-polarization means vanish, and the variances are given by ½(α²+β²+γ²), ½(α²−β²+γ²), and ½(α²+β²−γ²) for I, Q/U, and V respectively. These expressions lead to the spectrum-independent identities Var[Q]=Var[U], Var[1−I]=Var[Q]+Exp[V]², and Exp[1−I]²=Var[Q]+Var[V]. The paper also shows that a helical magnet","pith_inferences":["The spectrum-independent relations could be turned into a null test for claims of axion-induced polarization from extragalactic sources: if the measured ensemble statistics do not satisfy these identities, the source of the signal lies elsewhere.","If real cosmological magnetic fields are intermittent or shock-dominated, their non-Gaussianity would generically break the closed-form variance formulas; comparing observed variances with the Gaussian prediction could therefore diagnose field statistics as well as axion physics.","The single-domain assumption is likely the main obstacle to direct cosmological application; a multi-domain generalization would effectively replace the model spectra here with a direction-dependent superposition, possibly smearing the distinctive peak structures.","The helicity-dependent circular polarization may provide a way to measure the handedness of primordial magnetic fields using gamma-ray observations, independent of Faraday rotation measurements."],"forward_implications":["Observations of many equivalent photon sources behind the same magnetized region should show Var[Q] = Var[U], providing a direct test that requires no knowledge of the magnetic-field spectrum.","A statistically helical magnetic field generates a circular-polarization signal whose mean is proportional to the helical power spectrum, with a peak whose location in frequency encodes the axion mass or the field correlation length.","In the frequency window where the dimensionless combination Π(ω)d lies between O(1) and k_*d, the variance of the conversion probability is suppressed relative to its expectation value, making this window advantageous for axion searches.","The three consistency relations give a falsifiable set of equalities that any Gaussian-Born axion-photon conversion signal must satisfy; deviations would point to non-Gaussian magnetic fields, stronger coupling, or additional physics.","Because the same integral kernels appear in graviton-photon conversion, the statistical structure derived here transfers to other weakly coupled particle-mixing scenarios."],"fun_headline_variants":["Unpolarized light gets polarized by axion conversion in stochastic B-fields","Helical B-fields create circular polarization from axion-photon conversion","Stochastic B-fields force polarization even from initially unpolarized light","Axion conversion in random fields yields simple relations for polarization variances","Helical magnetic fields produce circular polarization peak in axion conversion"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The entire closed-form statistical machinery rests on the assumptions that the magnetic field is a Gaussian random field and that the conversion is weak enough for the first-order Born approximation to hold; if the real field is strongly non-Gaussian or the coupling is not perturbative, the formulas and the consistency relations lose their validity.","fun_headline_variants_meta":{"raw":{"variants":["Unpolarized light gets polarized by axion conversion in stochastic B-fields","Helical B-fields create circular polarization from axion-photon conversion","Stochastic B-fields force polarization even from initially unpolarized light","Axion conversion in random fields yields simple relations for polarization variances","Helical magnetic fields produce circular polarization peak in axion conversion"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001038,"raw_usage":{"total_tokens":4197,"prompt_tokens":728,"completion_tokens":3469,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":472,"completion_tokens_details":{"reasoning_tokens":3376}},"tokens_in":472,"tokens_out":3469,"duration_ms":26591,"temperature":1.0,"reasoning_tokens":3376,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T14:12:23.964632+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a set of many observed photon sources whose lines of sight traverse the same statistical magnetic-field region, measure their Stokes parameters, and compute the ensemble variance of Q and U; if the variances are unequal, or if the combination Var[1−I] − Var[Q] − Exp[V]² is measurably nonzero, the Gaussian-Born prediction is ruled out. Alternatively, a direct measurement of the four-point magnetic-field correlator would reveal non-Gaussianity and invalidate the Wick factorization at the heart of the derivation.","supporting_citations":[],"review_version":1}