{"id":"1683ba7b-cf2e-42ca-87c2-34ff186dc818","arxiv_id":"2504.13154","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"The combined analysis of SPTpol, ACT, POLARBEAR, and BICEP data constrains the anisotropic cosmic birefringence amplitude to ACB = 0.42^{+0.40}_{-0.34} × 10^{-4}, consistent with zero.","lead":"This paper combines CMB B-mode polarization data from four experiments to place a new upper limit on anisotropic cosmic birefringence, a rotation of light's polarization direction by parity-violating physics. The result, consistent with no signal, is the leading constraint on this effect and helps bound axion-like particle theories.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Reported 95% upper bound inconsistent with Table I: abstract says A_CB < 1e-4, full-dataset row gives 1.08e-4; text and table also conflict for the SPTpol-excluded case.","rationale":"The paper's main contribution is an upper bound on the amplitude of anisotropic cosmic birefringence. If the quoted 95% bound is not the one actually obtained, the abstract is misleading. The reader's weakest_assumption (model dependence) is real but explicitly acknowledged and does not cause an internal contradiction. The inconsistency in the upper limit is a concrete, verifiable error that directly affects the headline. The secondary issue of Eq. (16) is also worth correcting, but it affects only the robustness check in Sec. III.A; the main constraints in Table I do not include the isotropic term. Therefore the numeric upper-bound inconsistency is the most load-bearing. The verdict should remain conditional: the analysis is likely sound, but the paper must fix the reported bound and clarify the table/text mismatch before the claim as written is accepted. I partially agree with the reader because the reader mentioned the inconsistency in passing but chose a different weakest assumption.","tokens_in":11823,"tokens_out":9480,"duration_ms":77456,"concrete_test":"Take the MCMC chains (or rerun the likelihood) for the full combined dataset and for ACT+POLARBEAR+BICEP only, and compute the 95% credible upper limit from the marginalized posterior for A_CB. Compare the resulting values to Table I: if the full-data limit is 1.08 × 10^{-4} and the SPTpol-excluded limit is 0.85 × 10^{-4}, then the abstract's '<1 × 10^{-4}' and the text's '1.00 × 10^{-4}' are wrong and must be corrected. If the table values are wrong (e.g., due to a different quantization convention), the table must be corrected. Either way, the reported number must be made consistent with the posterior.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The abstract's headline upper bound 'A_CB < 1 × 10^{-4}' is not supported by the paper's own Table I. The full-dataset (SPTpol+ACT+POLARBEAR+BICEP) row reports a 95% CL upper limit of 1.08 × 10^{-4}, while the abstract quotes a bound that is 8% lower. In addition, the text in Sec. III states that the dataset combination excluding SPTpol yields a 95% CL upper limit of A_CB < 1.00 × 10^{-4}, but Table I's ACT+POLARBEAR+BICEP row gives 0.85 × 10^{-4} for the same combination. Because the paper's central claim is a constraint on A_CB, the numerical value of the upper bound is load-bearing; an unsupported or self-contradictory bound undermines the headline result even if the underlying likelihood is sound. This is an internal inconsistency, not a scope limitation, and it must be resolved before the abstract can be accepted as accurate.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper constrains the amplitude A_CB of anisotropic cosmic birefringence using B-mode polarization bandpowers from SPTpol, ACT, POLARBEAR, and BICEP, adopting the exact treatment of Namikawa (TN24) that goes beyond the thin last-scattering-surface approximation. The model assumes a massless pseudoscalar field with a scale-invariant primordial power spectrum, with the rotation-angle power spectrum scaling as C_αα ∝ A_CB/[L(L+1)]. The main results are a best-fit amplitude A_CB = 0.42^{+0.40}_{-0.34} × 10^{-4} for the full dataset combination, a 95% confidence-level upper limit quoted as A_CB < 1 × 10^{-4} in the abstract, and a robustness study that includes an isotropic rotation angle α. The paper concludes that the joint data show no significant detection of anisotropic cosmic birefringence and that the bounds are not dominated by a single experiment.","tokens_in":12103,"tokens_out":5140,"duration_ms":46898,"significance":"If the reported constraints are correct, this would be one of the leading CMB B-mode limits on anisotropic cosmic birefringence under the exact beyond-thin-LSS treatment, and the combination of four independent experiments is a useful step beyond previous single-experiment analyses. The paper also makes its analysis code publicly available at https://github.com/antolonappan/bbCAB, which is a clear strength. The main scientific conclusion—that current B-mode data do not prefer a nonzero anisotropic birefringence amplitude—is plausible and consistent with previous upper limits. However, the numerical discrepancies between the abstract, the text, and Table I directly affect the headline bound, and the isotropic-rotation template in Eq. (16) is missing a square, both of which must be corrected before the quoted limits can be considered reliable.","major_comments":[{"comment":"The abstract's headline 95% upper limit A_CB < 1 × 10^{-4} is not supported by Table I, where the full-dataset (SPTpol+ACT+POLARBEAR+BICEP) row reports a 95% upper limit of 1.08 × 10^{-4}. Similarly, Sec. III states that the dataset combination excluding SPTpol gives A_CB < 1.00 × 10^{-4}, but Table I lists 0.85 × 10^{-4} for the ACT+POLARBEAR+BICEP row. Because the paper's central claim is a numerical constraint on A_CB, this internal inconsistency is load-bearing and must be resolved—either by correcting the abstract and text to match Table I, or by recomputing the limits if the table is wrong.","section":"Abstract; Sec. III; Table I"},{"comment":"The isotropic rotation template is written as D^{iso,CB}_ℓ(α) = sin(2α) C^{EE}_ℓ, but the standard relation for rotation of Stokes parameters is C^{BB} = sin^2(2α) C^{EE} (for a pure E-mode input), so the equation is missing a square on the sine factor. This template is used in Sec. III.A and Fig. 3 for the joint fit with α, and the robustness claim 'A_CB < 1.0 × 10^{-4}' from that analysis is reported in the figure caption. The quoted joint-fit bound should not be trusted until Eq. (16) is corrected and the fit is repeated.","section":"Sec. III, Eq. (16)"}],"minor_comments":[{"comment":"The abstract says the result is 'consistent with zero within 2σ', but Table I reports 1.10σ for the full combination; this is technically true but vague, and the text should state the actual significance.","section":"Sec. III; Sec. IV"},{"comment":"Equation (15) includes the isotropic term D^{iso,CB}_ℓ, but the next paragraph says 'we did not include the isotropic term... in our model' for the main constraints; please clarify that Eq. (15) is the general model and that the baseline analysis is a restricted version with α fixed to zero.","section":"Sec. III, Eq. (15)"},{"comment":"The joint-fit 95% upper limit A_CB < 1.0 × 10^{-4} is quoted in the text and figure caption, but no corresponding row is given in Table I; the relationship between the joint-fit bound and the Table I bounds (e.g., why the joint limit is below the Table I full-dataset limit of 1.08 × 10^{-4}) should be explained.","section":"Sec. III.A; Fig. 3"},{"comment":"Several references are incomplete, lacking volume/page numbers (e.g., refs. [13], [16]-[18], [23], [44]); these should be completed before publication.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The numerical inconsistencies between the abstract, Sec. III, and Table I suggest that the manuscript may have been updated without propagating all changes; the editor should ensure the final version is internally consistent. The Eq. (16) missing-square issue is straightforward to fix but affects the robustness analysis. No concerns about novelty or scope."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The new content here is the combination of four B-mode datasets (SPTpol, ACT, POLARBEAR, BICEP) against Namikawa's exact anisotropic-birefringence template, plus a joint fit with isotropic rotation. That is a legitimate extension, not a new framework, and the main result—a null constraint at roughly ACB < 1e-4—is plausibly the leading B-mode limit on this amplitude. The analysis is described clearly, the likelihood is standard Gaussian, and the code is public. Credit where it's due: this is a reproducible, useful constraint, and the SPTpol-only 1.8σ preference weakening to 1.1σ in the full combination is a sensible and honestly reported finding.\n\nThe soft spots are real, and they are exactly where the stress-test note lands. The abstract says ACB < 1e-4, but Table I's full-combination row gives 1.08e-4. That's not rounding; it's an 8% discrepancy in the paper's headline number. The text also says the no-SPTpol combination gives a 95% upper limit of 1.00e-4, while Table I lists 0.85e-4 for that same row. The paper contradicts its own table twice, and since the constraint value is the point of the paper, the abstract cannot be accepted as written. Minor but not trivial: Eq. (16) writes the isotropic-rotation B-mode as sin(2α) C_EE^ℓ; the standard expression is sin^2(2α) C_EE^ℓ. That typo affects the joint fit in Sec. III.A, the very analysis the abstract cites for robustness under isotropic rotation. The primary ACB constraints in Table I do not use this equation, so the main null result survives, but the robustness claim is weakened until the formula is corrected and the joint fit rechecked.\n\nThe limitations section is honest: the constraint applies to a massless, scale-invariant pseudoscalar, and the paper does not test other spectral shapes. That is a scope statement, not a flaw, given the template is borrowed from TN24. The citation pattern is fine; self-citations are tangential.\n\nWho is this for? Anyone working on CMB polarization and axion-like photon couplings. It deserves a serious referee, but the referee should require the abstract/table consistency fix and the Eq. (16) correction before publication. I would send it to peer review, and I'd cite the corrected version as a leading B-mode constraint. Take it to reading group once the authors clean up the numbers.","headline":"Useful multi-experiment null constraint on anisotropic cosmic birefringence, but the headline upper bound doesn't match the paper's own Table I and the isotropic-rotation template in Eq. (16) is missing a square; both need fixing before the abstract is accurate.","tokens_in":12560,"tokens_out":2184,"would_cite":true,"duration_ms":20683,"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":"Four CMB surveys combine to place the tightest B-mode bound on anisotropic cosmic birefringence.","keywords":["cosmic birefringence","anisotropic birefringence","CMB B-mode polarization","axion-like particles","parity violation","last-scattering surface","polarization rotation","CMB experiments"],"falsifier":"A future high-sensitivity CMB experiment that pushes the 95% upper limit below roughly $3\\times10^{-5}$ would test the point directly: if the best-fit $A_{\\rm CB}$ rises above $1\\times10^{-4}$ with $5\\sigma$ significance, the paper's null-consistent conclusion is wrong. Alternatively, a direct quadratic-estimator measurement of the rotation-angle power spectrum whose shape departs from $1/[L(L+1)]$ would falsify the scale-invariant model under which the bound is derived.","tokens_in":11681,"feed_emoji":"🔭","tokens_out":11114,"duration_ms":92563,"temperature":0.7,"pith_summary":"The paper asks whether the rotation of CMB polarization by an anisotropic pseudoscalar field—the anisotropic part of cosmic birefringence—can be seen in B-mode polarization, the curl component of the microwave sky, using a treatment that does not approximate the last-scattering surface as infinitely thin. Combining B-mode spectra from SPTpol, ACT, POLARBEAR, and BICEP, it finds a best-fit amplitude $A_{\\rm CB} = 0.42^{+0.40}_{-0.34}\\times 10^{-4}$, consistent with zero within $2\\sigma$, and a 95% upper limit $A_{\\rm CB} < 1\\times 10^{-4}$. The SPTpol-only hint of a $1.8\\sigma$ preference weakens as the other datasets are added, and allowing an isotropic rotation angle further pulls the anisotropic amplitude toward zero. If the result holds, it is the current tightest B-mode constraint on anisotropic cosmic birefringence under the exact, beyond-thin-LSS calculation.","feed_headline":"Four CMB surveys bound anisotropic birefringence below 10^-4","feed_subtitle":"Joint B-mode data from SPTpol, ACT, POLARBEAR, and BICEP stay consistent with zero rotation at 2 sigma.","key_machinery":"The central object is the dimensionless amplitude $A_{\\rm CB} = (g_\\phi/2)^2 (H_I/2\\pi)^2$, defined so that the rotation-angle power spectrum is $C^{\\alpha\\alpha}_L \\simeq 2\\pi/[L(L+1)]\\,A_{\\rm CB}$ at large angular scales. The computation carries this amplitude through the exact B-mode formula of Ref. [53], which uses the total angular momentum method to integrate the polarized radiative-transfer equation over the finite visibility function rather than freezing the rotation at a single last-scattering epoch. The key intermediate is the distorted E-mode spectrum $C^{EE}_{\\ell',L}$ built from the projected polarization source $s_\\ell(q,\\eta)$ and the birefringence kernel $u_L(k,\\eta)=(g_\\phi/2)\\,j_L(k(\\eta_0-\\eta))T(k,\\eta)$, whose product sets how much E-polarization is converted into B-modes. In this treatment the predicted B-mode spectrum is suppressed by roughly an order of magnitude at $\\ell\\lesssim10$ and a factor of two at $\\ell\\gtrsim100$ relative to the thin-LSS approximation, which is why the exact template matters for the quoted bound.","core_discovery":"On its own terms, the paper's central claim is that the anisotropic-birefringence B-mode signal, computed with the finite thickness of the last-scattering surface included, is not detected in the combined data: the joint likelihood prefers $A_{\\rm CB} = 0.42^{+0.40}_{-0.34}\\times 10^{-4}$ (about $1.1\\sigma$), and the 95% confidence interval excludes amplitudes above $1\\times 10^{-4}$. Two robustness results accompany the main bound: no single experiment drives the constraint, and marginalizing over an isotropic rotation angle $\\alpha$ broadens the posterior and moves the best fit from about $4.6\\times 10^{-5}$ to $1.9\\times 10^{-5}$, so the anisotropic signal does not survive as a significant detection. The authors present the bound as the leading constraint of its kind, derived under a massless, scale-invariant pseudoscalar model.","pith_inferences":["The quoted bound is tied to a massless, scale-invariant pseudoscalar model; a massive axion or a defect network would imprint a different $L$-dependence in $C^{\\alpha\\alpha}_L$, so the same data would need a separate spectral-shape analysis before any limit is quoted.","A direct tomographic reconstruction of the rotation-angle power spectrum from EB correlations, rather than a template fit to the B-mode auto-spectrum, could break the $A_{\\rm CB}$–$\\alpha$ degeneracy and test the assumed spectral shape.","Because the exact treatment suppresses large-scale B-modes, the quickest route to beating the $1\\times10^{-4}$ bound may be small-scale surveys with aggressive delensing rather than low-multipole observations alone."],"forward_implications":["The 95% upper limit $A_{\\rm CB} < 1\\times 10^{-4}$ becomes the reference B-mode bound for anisotropic cosmic birefringence under the exact finite-LSS treatment, replacing earlier estimates made with the thin-LSS approximation.","The SPTpol-only $1.8\\sigma$ preference weakens when ACT, POLARBEAR, and BICEP are added, so the combined result is the one to use in future model comparisons.","Marginalizing over an isotropic rotation angle shifts the anisotropic best fit from about $4.6\\times10^{-5}$ to $1.9\\times10^{-5}$, showing that future analyses must fit both components jointly to avoid overestimating the anisotropic signal.","Because the exact treatment suppresses the B-mode signal at low multipoles by roughly an order of magnitude, large-angle B-mode surveys face a higher bar for detecting this effect than thin-LSS templates suggested."],"supporting_citations":[{"why":"Supplies the exact beyond-thin-LSS B-mode power spectrum and the numerical framework the likelihood adopts.","marker":"[53]"},{"why":"Provides the ACT DR6 B-mode bandpowers used in the combined fit.","marker":"[23]"},{"why":"Provides the BICEP/Keck B-mode measurements that anchor low-multipole sensitivity.","marker":"[64]"},{"why":"Provides the POLARBEAR degree-scale B-mode data included in the joint likelihood.","marker":"[65]"},{"why":"Provides the SPTpol B-mode spectra, foreground templates, and calibration and beam priors that define the baseline model.","marker":"[66]"},{"why":"Defines the rotation-angle power spectrum $C^{\\alpha\\alpha}_L$ and the $A_{\\rm CB}$ parametrization used in the templates.","marker":"[61]"},{"why":"Supplies the total angular momentum formalism underlying the exact mode-coupling calculation.","marker":"[59]"}],"fun_headline_variants":["Anisotropic birefringence stays consistent with zero in B-modes","Four CMB experiments rule out large anisotropic birefringence","CMB B-mode data cap anisotropic birefringence at 1e-4","No anisotropic birefringence detected across four surveys","Joint CMB polarization analysis tightens birefringence bound"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"Everything rests on assuming the anisotropic rotation is produced by a massless pseudoscalar field whose primordial fluctuations are scale-invariant, so the rotation-angle power spectrum has the shape $C^{\\alpha\\alpha}_L\\propto 1/[L(L+1)]$; if the true signal has a different multipole dependence, the quoted $A_{\\rm CB}$ bound does not directly apply.","fun_headline_variants_meta":{"raw":{"variants":["Anisotropic birefringence stays consistent with zero in B-modes","Four CMB experiments rule out large anisotropic birefringence","CMB B-mode data cap anisotropic birefringence at 1e-4","No anisotropic birefringence detected across four surveys","Joint CMB polarization analysis tightens birefringence bound"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000291,"raw_usage":{"total_tokens":1721,"prompt_tokens":988,"completion_tokens":733,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":604,"completion_tokens_details":{"reasoning_tokens":636}},"tokens_in":604,"tokens_out":733,"duration_ms":7115,"temperature":1.0,"reasoning_tokens":636,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T12:13:47.355923+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A future high-sensitivity CMB experiment that pushes the 95% upper limit below roughly $3\\times10^{-5}$ would test the point directly: if the best-fit $A_{\\rm CB}$ rises above $1\\times10^{-4}$ with $5\\sigma$ significance, the paper's null-consistent conclusion is wrong. Alternatively, a direct quadratic-estimator measurement of the rotation-angle power spectrum whose shape departs from $1/[L(L+1)]$ would falsify the scale-invariant model under which the bound is derived.","supporting_citations":[],"review_version":1}