{"id":"d96bb005-0c4b-40d3-91cb-b8b78e76a736","arxiv_id":"2506.07546","paper_version":1,"verdict":"REJECT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"high","formal_verification":"none","parameter_count":4,"one_line_summary":"The paper derives a first bound on the dimensionless axion-photon coupling gγ from pulsar polarization data, reporting |gγ|<0.93 at 1σ for axion masses below 10^-11 eV, via a neutron-star-induced axion field.","lead":"Physicists claim that the polarization of radio pulses from the pulsar PSR B1919+21, measured by the FAST telescope, can be used to constrain the axion-photon coupling gγ independently of the axion decay constant. They report a first, weak limit, |gγ|<0.93 at 1σ, for ultra-light axions, but the analysis depends on several untested assumptions about neutron stars and pulsar emission.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Fit cannot separate g_gamma from the free pulsar factor A: for m_a < 10^-11 eV the signal is Delta(alpha) = -6.59 g_gamma A (omega/GHz)^2, so the quoted |g_gamma| < 0.93 requires an unstated prior on A.","rationale":"The reader correctly identifies the frequency-independent intrinsic PPA assumption as a serious astrophysical weakness, and the reader's rationale also mentions the coupling-pulsar degeneracy. I would place the primary weight on the degeneracy, because it is an internal identifiability failure that does not depend on pulsar emission modeling. The abstract claims a constraint that 'uniquely depends' on g_gamma, yet the fit treats A as a free parameter; for the mass range in which the constraint is quoted, the signal is exactly the product g_gamma A. The paper does not report a prior on gamma or A, so the quoted 1-sigma bound on g_gamma is not derivable from the stated procedure. This is a missing-identification problem, not a disagreement with external consensus. The 3-sigma value also differs between the constraints section (1.73) and the conclusion (1.93), and the neglected RM systematic is explicitly admitted; these support the same rejection. The paper's method may become viable if A is independently constrained, but the central claim as stated is not supported.","tokens_in":8560,"tokens_out":12835,"duration_ms":172235,"concrete_test":"Analytical check: substitute Eq. (11) into Eq. (14) and show that for m_a -> 0 the chi-squared depends on g_gamma and A only through B = -6.59 g_gamma A / GHz^2, so the Fisher matrix has a zero eigenvalue along g_gamma d/dg_gamma - A d/dA. Then rerun the fit with A fixed to the value implied by the stated gamma prior, e.g. gamma = 100 gives A = (R_NS / 10 km)(PNS / 1 s), and compare the 1-sigma interval. If the paper used an implicit prior on gamma or A, it must be stated; if no finite g_gamma interval survives without such a prior, the central constraint is unsupported.","verdict_should_be":"REJECT","load_bearing_attack":"The most load-bearing problem is not primarily astrophysical systematics but identifiability. Equation (11) gives Delta(alpha) = -6.59 g_gamma A (omega/GHz)^2 for m_a < 10^-11 eV. The text states that the fit has four free parameters: g_gamma, m_a, pulsar factor A, and initial angle alpha_0. In the m_a -> 0 limit, the likelihood depends on g_gamma and A only through the product B = -6.59 g_gamma A / GHz^2: Q = cos(B omega^2 + alpha_0), U = sin(B omega^2 + alpha_0). Therefore the transformation (g_gamma, A) -> (g_gamma / s, s A) leaves all predictions unchanged. With A free and no reported prior on gamma, and hence on A, the profile likelihood in g_gamma is flat along this degeneracy, and no finite 1-sigma interval for |g_gamma| follows from the fit. The paper also states that 'The possible systematic errors caused by RM uncertainty is not included,' which weakens the claimed precision, but the degeneracy is more fundamental: the headline constraint is empty unless an independent constraint on A is imposed.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper proposes that in neutron stars, restoration of chiral symmetry shifts the QCD axion field to a VEV of order πf_a, with a Yukawa-like profile outside the star. Using radius-frequency mapping to assign a radial emission height to each radio frequency, the authors convert the axion-induced birefringence into a predicted frequency-dependent polarization rotation that is formally independent of f_a. They fit this model to FAST polarization data of PSR B1919+21 and report |g_γ|<0.93 at 1σ for m_a<10^-11 eV, with weaker constraints at higher mass. The paper includes a derivation of the axion profile from a modified UV potential (Scenario II) and a derivation of the radius-frequency mapping.","tokens_in":8834,"tokens_out":6995,"duration_ms":80972,"significance":"The idea of using an f_a-independent axion VEV in neutron stars to probe the quantized coupling g_γ is original and, if the underlying scenario is realized, would be an interesting complement to laboratory searches. The authors use real FAST observations and are transparent about the omission of RM systematics. However, the headline constraint is not supported by the analysis as written: in the massless limit the observable depends only on the product g_γ A, with A a free parameter, so the reported 1σ bound on g_γ is derived only with an unstated prior. Together with the model-dependence of the axion profile and the untested assumption of frequency-independent intrinsic polarization angles, this makes the current claim too fragile to be published as a constraint.","major_comments":[{"comment":"In the limit m_a < 10^-11 eV, Eq. (11) gives Δα(ω_c) = -6.59 g_γ A (ω_c/GHz)^2. The fit described in the same section treats the pulsar factor A as a free parameter (along with g_γ, m_a, and α_0). For m_a below this threshold the likelihood depends on g_γ and A only through the product g_γ A, since Q_th and U_th in Eqs. (12)-(13) are cos/sin of that product plus α_0. The transformation (g_γ, A) -> (g_γ/s, s A) leaves every prediction unchanged, so the profile likelihood in g_γ is flat and no finite 1σ interval |g_γ| < 0.93 follows from the fit without an external prior on A (or on γ, which enters A through the sixth power). The authors must impose and justify an independent constraint on A, or quote the constraint as one on the product g_γ A rather than on g_γ.","section":"Constraints from FAST data, Eq. (11)"},{"comment":"The paper assumes 'the frequency-independent nature of initial polarization angles [52]' and attributes all observed frequency dependence of the polarization position angle to the axion. This is not established for pulsars: radius-to-frequency mapping itself implies emission at different heights for different frequencies, and magnetospheric propagation (including refraction and mode coupling) produces frequency-dependent position angles independent of any axion effect. Since the predicted axion signal is purely frequency dependent, this assumption is load-bearing. A quantitative assessment of the intrinsic frequency dependence for PSR B1919+21 (or a multi-frequency control with a different pulsar) is needed before a signal can be attributed to axion birefringence.","section":"Axion detection, paragraph before Eq. (5)"},{"comment":"The derived neutron-star profile and the subsequent bound are contingent on a very specific UV potential: β1=0 with parameters satisfying inequality (21), which permits the vacuum to shift to a≈πf_a when ⟨qq⟩→0. No concrete UV completion realizing this choice is given, and the claim in the abstract to derive 'the first constraint' from neutron stars is therefore conditional on a non-generic model assumption. The authors should either provide an explicit model satisfying Eq. (21) or soften the claim to a constraint within Scenario II. A useful additional check would be to show how the constraint changes as β1 and the ratio m_q⟨qq⟩/(√2 Λ^3 f_a |λ1|) are varied within the allowed region.","section":"Axion phase transition in neutron stars, Scenario II, Eq. (21)"},{"comment":"The text states that 'the possible systematic errors caused by RM uncertainty is not included,' even though an RM uncertainty of ±1.1 rad m^-2 corresponds to roughly 3° of polarization angle across the 1-1.5 GHz band. For a signal whose amplitude is to be constrained at the order of a radian, a 3° (≈0.05 rad) unmodelled uncertainty is not negligible and should be marginalised over or added in quadrature. The authors should quantify how the reported contours change when the RM is varied by its uncertainty.","section":"Constraints from FAST data, after Faraday rotation"}],"minor_comments":[{"comment":"The 3σ limit quoted in the Conclusion (|g_γ|<1.93) differs from the value 1.73 given in the main text; please reconcile these numbers.","section":"Conclusion vs. Constraints section"},{"comment":"The subscript in g_γ is rendered with a space in the title and abstract; fix the typographical rendering.","section":"Title and abstract"},{"comment":"Reference [52] is a textbook; please provide a primary reference or a specific chapter/table for the claimed frequency independence of initial polarization angles.","section":"Reference [52]"},{"comment":"Since A is defined by Eq. (8) in terms of R_NS, P_NS, and γ but is treated as a free parameter, the relationship between the fitted A and the nominal pulsar parameters should be discussed; in particular, the value of γ implied by the best-fit A should be compared with the expected range γ ~ 10^2-10^3 cited in the Appendix.","section":"Eq. (8) and fitted A"}],"recommendation":"reject","confidential_remarks":"The central statistical degeneracy between g_γ and A is the decisive issue; I do not think a revision within the manuscript's scope can fix it without a fundamentally different analysis. If the authors supply an independent, well-justified prior on A or γ, the paper could be reconsidered."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The new thing here is real: the neutron-star axion VEV with the π f_a profile, combined with radius-frequency mapping, gives an observable that is formally independent of f_a. That is a clever move and worth knowing about if you work on axion phenomenology or generalized symmetries. The paper also does something rare: it goes from the field-theory argument to an actual FAST data set and a quoted number. The algebra is transparent, and the f_a cancellation is a legitimate consequence of the assumed profile, not a circular fit. The citation pattern looks fine; the building blocks are from the right prior literature, and the self-citation to the companion data-processing paper is legitimate given they did that analysis.\n\nBut the central constraint does not survive contact with the fitting procedure. For m_a < 10^{-11} eV, Eq. (11) is Δα = -6.59 gγ A (ω/GHz)^2. The likelihood then depends on gγ and A only through the product. The transformation (gγ, A) -> (gγ/s, sA) leaves every prediction unchanged. With A a free parameter and no prior stated anywhere, the profile likelihood in gγ is flat along that degeneracy. The reported |gγ| < 0.93 at 1σ is not a statement about gγ alone; it is an artifact of however the fit happened to explore the product. The paper explicitly lists four free parameters and never constrains A independently, so the headline bound is empty unless an unstated prior on A is imposed.\n\nThere are other soft spots, in proportion. The astrophysical chain — complete chiral restoration inside the star, Scenario II of the UV potential, the piecewise profile with a kink at the surface, and especially the frequency-independent intrinsic polarization angle — is a stack of assumptions, and pulsar magnetospheres are known to produce frequency-dependent position-angle structure. The paper also concedes RM systematic errors are not included. And the 3σ limit changes from 1.73 in the results section to 1.93 in the conclusion, which is sloppy and erodes trust.\n\nNone of this kills the idea. The mechanism is interesting enough that I would not desk-reject it. But the claimed constraint is not supported as stated. A serious referee should insist on fixing the degeneracy first: impose a physically motivated prior on the pulsar factor A, or fix it from independent emission modeling, then report the profile likelihood along the degenerate direction. The manuscript also needs to reconcile the 3σ numbers and at least bound the RM systematics.\n\nWho is this for? People working on axion detection beyond the standard f_a-suppressed probes will find the mechanism stimulating, but they should not walk away with the number.\n\nRecommendation: send to peer review, but expect a heavy revision. The method deserves referee time; the current bound does not deserve publication as is.","headline":"Clever new mechanism for probing gγ without f_a, but the claimed bound is empty because gγ and the pulsar factor A are degenerate in the fit.","tokens_in":9395,"tokens_out":4678,"would_cite":false,"duration_ms":53297,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["14.80.Va","95.35.+d","97.60.Gb"],"model":"deepseek-v4-flash","headline":"This paper proposes that a neutron star's restored chiral symmetry sets the axion field to a large VEV whose Yukawa tail induces a frequency-dependent radio polarization rotation, giving the first f_a-independent constraint on the…","keywords":["axion-photon coupling","axion decay constant","neutron star","pulsar polarization","birefringence","radius-to-frequency mapping","PSR B1919+21","FAST telescope"],"falsifier":"Resolve the polarization angle by pulse phase: geometric (rotating-vector) effects create a phase sweep that shifts with frequency, whereas the axion rotation is phase-independent; observing such a frequency-shifting sweep would rule out the axion interpretation. A second check: compare the best-fit A against independent estimates of γ and R_NS; if γ must sit far outside $10^{2}$–$10^{3}$, the model is absorbing non-axion physics.","tokens_in":8352,"feed_emoji":"📡","tokens_out":12404,"duration_ms":118818,"temperature":0.7,"pith_summary":"Neutron stars are dense enough to restore QCD chiral symmetry, and the paper argues that this forces the axion field inside the star to a large value π f_a rather than zero. Outside the star the field falls off in a Yukawa tail, and radio photons crossing this region have their polarization rotated by an angle that depends on the dimensionless axion-photon coupling g_γ only — the decay constant f_a cancels out. Using 1–1.5 GHz FAST observations of PSR B1919+21 and a radius-to-frequency mapping that ties emission altitude to wavelength, the authors fit the observed Stokes parameters and obtain the first constraint |g_γ| < 0.93 at 1σ (1.33 at 2σ) for axion masses below $10^{-11}$ eV. The authors note that systematic errors from the rotation-measure uncertainty are not included in these quoted confidence levels. The result matters because it opens a way to probe the quantized coupling g_γ, a quantity linked to the global structure of the Standard Model, without relying on large f_a suppression or exotic cosmological relics.","feed_headline":"Neutron star data cap axion-photon coupling below 0.93","feed_subtitle":"Radio polarization of PSR B1919+21 yields the first axion-photon bound independent of the decay constant.","key_machinery":"The load-bearing object is the macroscopic axion field profile a(r) around a neutron star, with the VEV π f_a inside and a Yukawa tail outside. The key identity is the cancellation: because the field amplitude scales as f_a, the ratio (a_f − a_i)/(2f_a) in the birefringence formula becomes independent of f_a and leaves only g_γ. The second essential piece is the radius-frequency mapping ω_c ∝ $γ^{3}$ $P_NS^{{-1/2}}$ $r^{{-1/2}}$, which converts the radial profile into a frequency-dependent rotation that can be read off a single pulsar's radio spectrum.","core_discovery":"The central discovery is that a neutron star's axion field profile — π f_a inside the star, π f_a (R_NS/r) $e^{{-m_a(r-R_NS)}}$ outside — combined with the photon birefringence formula Δα = (g_γ/2f_a)(a_f − a_i), produces a rotation angle Δα(ω) = −6.59 g_γ A (ω/GHz)^2 exp[−m_a R_NS(0.24 A (GHz/ω)^2 − 1)] in which every f_a dependence has cancelled. The parameter A = (R_NS/10 km)(P_NS/1 s)(100/γ)^6 encodes the star's radius, spin period, and magnetospheric Lorentz factor. For ultralight axions the exponential is negligible and the rotation is simply ∝ g_γ A ω². Fitting this law to FAST data on PSR B1919+21 with g_γ, m_a, A, and the initial polarization angle α_0 as free parameters yields |g_γ| < 0.93 at 1σ for m_a < $10^{-11}$ eV, and sensitivity degrades for m_a > $10^{-10}$ eV because of the exponential suppression.","pith_inferences":["The cancellation that makes $g_\\gamma$ observable relies on the axion settling at exactly $\\pi f_a$ inside the star; any equation-of-state effect that shifts this value would re-introduce $f_a$ dependence and change the predicted normalization — a robustness test the paper does not run.","Pulsar polarization is known to have intrinsic frequency structure; a natural next check is to fit the same data with a slowly varying polynomial background and see whether the $\\omega^2$ axion term survives — this would separate the axion hypothesis from radius-to-frequency mapping artifacts.","Because the signal scales as $A \\propto \\gamma^{-6}$, the method is most sensitive to old, slow pulsars with small Lorentz factors; targeting such objects could improve the constraint far more than collecting more photons."],"forward_implications":["Axion searches no longer have to be suppressed by an unknown $f_a$: this method isolates the dimensionless coupling $g_\\gamma$ directly from a single star's radio spectrum.","For $m_a < 10^{-11}$ eV the bound is mass-independent, so a positive detection would be a clean $\\omega^2$ signature rather than a broad parameter-space fit.","The same technique applied to other pulsars with well-measured polarization would produce independent $g_\\gamma$ bounds, and combining them could push the limit below 0.9.","For axions heavier than about $10^{-10}$ eV the exponential tail suppresses the signal, so the method's reach is limited to ultra-light masses."],"supporting_citations":[{"why":"Supplies the radius-frequency mapping $\\omega_c \\propto \\gamma^3 P^{-1/2} r^{-1/2}$ that converts emission altitude into radio frequency.","marker":"[37]"},{"why":"Derives the axion field profile around a neutron star with the in-star VEV $\\pi f_a$, the central profile used in the rotation calculation.","marker":"[40]"},{"why":"Provides the claim that initial polarization angles are frequency-independent, the background assumption on which the fit's interpretation rests.","marker":"[52]"},{"why":"Establishes the photon birefringence formula $\\Delta\\alpha = g_\\gamma(a_f - a_i)/(2f_a)$ for axions.","marker":"[49]"},{"why":"The source of the FAST observational data for PSR B1919+21 used in the $\\chi^2$ fit.","marker":"[53]"},{"why":"Supplies the rotation measure RM = -13.1 ± 1.1 rad m^{-2} used to calibrate Faraday rotation across the band.","marker":"[60]"},{"why":"Provides the QCD axion potential used in the supplemental derivation of the phase transition and the $\\pi f_a$ VEV.","marker":"[61]"}],"fun_headline_variants":["Neutron star sets first axion-photon coupling bound","Radio pulsar data cap axion-photon coupling at 0.93","Axion-photon coupling limited by neutron star observation","PSR B1919+21 yields first axion-photon constraint","Neutron star probes axion coupling independent of decay constant"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole analysis assumes that a pulsar's intrinsic polarization angle at emission is the same at all radio frequencies, so any frequency dependence in the observed data is attributed to the axion; if the pulsar's magnetosphere itself produces frequency-dependent polarization angles, the constraint could be mimicking ordinary astrophysics.","fun_headline_variants_meta":{"raw":{"variants":["Neutron star sets first axion-photon coupling bound","Radio pulsar data cap axion-photon coupling at 0.93","Axion-photon coupling limited by neutron star observation","PSR B1919+21 yields first axion-photon constraint","Neutron star probes axion coupling independent of decay constant"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000699,"raw_usage":{"total_tokens":3131,"prompt_tokens":894,"completion_tokens":2237,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":510,"completion_tokens_details":{"reasoning_tokens":2148}},"tokens_in":510,"tokens_out":2237,"duration_ms":16065,"temperature":1.0,"reasoning_tokens":2148,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T05:33:27.843630+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Resolve the polarization angle by pulse phase: geometric (rotating-vector) effects create a phase sweep that shifts with frequency, whereas the axion rotation is phase-independent; observing such a frequency-shifting sweep would rule out the axion interpretation. A second check: compare the best-fit A against independent estimates of γ and R_NS; if γ must sit far outside $10^{2}$–$10^{3}$, the model is absorbing non-axion physics.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the radius-frequency mapping $\\omega_c \\propto \\gamma^3 P^{-1/2} r^{-1/2}$ that converts emission altitude into radio frequency."},{"cited_title":"Lyne and F","cited_arxiv_id":null,"evidence_quote":"Provides the claim that initial polarization angles are frequency-independent, the background assumption on which the fit's interpretation rests."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The source of the FAST observational data for PSR B1919+21 used in the $\\chi^2$ fit."},{"cited_title":"Rapid Rotation of Polarization Orientations in PSR B1919+21's Single Pulses: Implications On Pulsar's Magnetospheric Dynamics","cited_arxiv_id":"2411.18999","evidence_quote":"Supplies the rotation measure RM = -13.1 ± 1.1 rad m^{-2} used to calibrate Faraday rotation across the band."}],"review_version":1}