{"id":"621a125a-5db6-400b-886f-74e11634a913","arxiv_id":"2501.13662","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"An implementation of the Passaglia-Hu effective fluid approximation in CAMB achieves accurate CMB predictions for ultralight axions and yields new upper limits on their abundance from Planck PR4 and DESI BAO data.","lead":"This paper adds a fast, accurate way to compute how ultralight axion dark matter affects the cosmic microwave background inside a popular cosmology code, and uses it to tighten limits on axion mass and abundance. A smart generalist might care because ultralight axions are a leading dark matter candidate, and the new limits help narrow where to look.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Validation compares PH EFA at early switch against PH EFA at late switch, not against exact KG; the abstract's sub-percent accuracy claim rests on the untested assumption that the late-switch reference is exact.","rationale":"The central contribution is a fast implementation whose advertised accuracy is relative to exact KG solutions. The most load-bearing condition is therefore that the validation reference itself be accurate. The paper validates only against a later-switch version of the same PH EFA, so systematic errors in PH EFA are common to both model and reference and cancel in the χ2 comparison. This is the same weakness the reader identified as the weakest assumption. It is not fatal: the method is plausibly accurate, the code is public, and a direct exact-KG comparison is feasible for the relevant masses. The agreement with the independent implementation in Ref. [30] is reassuring for code correctness, but both codes implement the same PH EFA, so it does not test accuracy against exact KG. The conditional verdict already issued remains appropriate; the authors should either add the direct exact comparison or revise the abstract's accuracy claim. No change in verdict is needed.","tokens_in":12579,"tokens_out":5590,"duration_ms":52137,"concrete_test":"For m=10^-27 eV (the most sensitive case, with oscillations starting near matter-radiation equality), compute C_l^TT using AxiCAMB production settings (m/H*=50, accuracy=1). Then integrate the exact perturbed Klein-Gordon equations (Eqs. 5–8) for the same initial conditions and cosmology with no EFA switch, using a high-accuracy ODE solver through the oscillatory regime, and compute C_l^TT with the same CAMB settings. Report max |ΔC_l|/C_l over 2≤l≤2000. If this exceeds 1%, the Fig. 7 ground truth is not exact and the abstract's sub-percent claim is unsupported; if it stays below 1% for a set of representative masses (10^-28, 10^-27, 10^-26, 10^-25, 10^-24 eV), the concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's headline claim—'Compared to exact solutions of the Klein-Gordon equation, our method achieves sub-percent accuracy'—is not actually tested in the body. Section III.C adopts 'the PH EFA with a switch at m/H*=300 and accuracy=3' as the ground-truth reference for Fig. 7. This is the same approximation at a later switch, not an exact KG solution. The m/H*=1000 discussion in Figs. 3–4 likewise compares the two fluid prescriptions to each other, not to exact field evolution. The validation therefore cannot detect errors common to PH EFA at all switch times, including errors in the effective equation of state (Eq. 16), the sound speed (Eq. 26), the matching conditions (Eqs. 17–18 and 25), or the weighting/grace-period implementation. If the late-switch PH EFA is biased relative to exact KG by more than roughly 1% in C_l, the central accuracy claim fails and the MCMC limits inherit that bias. The cited Passaglia–Hu calibration is not reproduced here, and the note-added comparison with Ref. [30] checks two PH EFA codes against each other, which does not close the loop. The issue is addressable: a direct exact-KG comparison for the most sensitive masses near 10^-27–10^-28 eV would settle it.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper describes AxiCAMB, a public implementation of the Passaglia–Hu effective fluid approximation (PH EFA) for ultralight axions in the CAMB Boltzmann code. The authors derive the background and perturbation equations, detail the numerical switching between Klein–Gordon evolution and the fluid description, and show reduced sensitivity to the switch time compared to the standard EFA. They then perform MCMC analyses with Planck PR4 and DESI BAO data to obtain upper limits on the axion fraction and physical density for fixed masses from 10^-28 to 10^-24 eV, reporting improved constraints relative to previous standard-EFA analyses. The abstract claims sub-percent accuracy relative to exact Klein–Gordon solutions, and that claim is the focus of my concerns.","tokens_in":12845,"tokens_out":7587,"duration_ms":63444,"significance":"If the sub-percent accuracy claim is fully supported, AxiCAMB would be a valuable public tool for fast and accurate ULA parameter estimation, and the demonstration of switch-time robustness is a useful contribution. The code is open source, and the MCMC pipeline is described in sufficient detail for reproduction. However, the headline accuracy claim is not directly validated against exact Klein–Gordon solutions in the body of the paper; the validation instead compares the PH EFA with an early switch to the same approximation with a late switch. This leaves the cosmological constraints vulnerable to systematic errors common to the approximation, so the significance of the results depends on closing that validation gap.","major_comments":[{"comment":"The paper uses the phrase 'ground truth' inconsistently. In Section III.B, the m/H*=1000 case is called 'a ground-truth benchmark', while Section III.C adopts the PH EFA with m/H*=300 and accuracy=3 as 'a ground-truth reference'. Neither is an exact solution; both are the same approximation at different switch times. This is not merely a wording issue: it means the central accuracy claim is not tied to an external standard, and the choice of reference switch time can itself affect the inferred chi^2. The authors should state explicitly what the reference is, and justify why m/H*=300 is sufficient to serve as a proxy for the exact solution (e.g., by showing convergence of the CMB spectrum with increasing m/H* for the PH EFA).","section":"III.B and III.C"},{"comment":"The weighting function defined in Eq. (30), W(a)=exp(alpha(ln a* - ln a)) with alpha>0, is not bounded above; for a << a*, W(a) is larger than 1 and the combination rho_ax = W rho_KG + (1-W)rho_fluid involves a negative weight on the fluid density. The accompanying text states that 'For a << a*, the evolution is dominated by the exact field solution', which is not what this formula implies unless rho_fluid is negligible in that regime. Please clarify the domain of validity of Eq. (30) and either use a bounded weighting (e.g., a sigmoid or a normalized average) or justify the unbounded form and show that it does not introduce artifacts in the background and perturbation evolution.","section":"III.A, Eq. (30)"}],"minor_comments":[{"comment":"There is a typo in the sentence 'the the full KG evolution' (should be 'the full KG evolution'), and 'reyonization' should be 'reionization' in the same section.","section":"III.B"},{"comment":"The caption reads 'As for Fig. 6 but for a switch m/H*=10.' It should refer to Fig. 5, not Fig. 6.","section":"Fig. 6 caption"},{"comment":"The convergence criterion states 'we require R1 <= 0.05', which is inconsistent with the standard Gelman-Rubin statistic; this should likely read 'R-1 <= 0.05' (or equivalently R <= 1.05). Please clarify to ensure reproducibility.","section":"IV"},{"comment":"There are several small textual issues: 'max' is used instead of 'm_ax' in the Introduction and elsewhere, and the name 'Ureña-López' is rendered incorrectly in the bibliography. These should be corrected in a final polish.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The main blocker is the validation: the sub-percent accuracy claim relative to exact Klein-Gordon solutions is not demonstrated, because the reference is the same approximation at a later switch. This is fixable with a direct comparison to exact KG for a subset of modes and masses, and the rest of the paper appears sound. If the authors provide that comparison, the paper would likely be acceptable. The MCMC constraints are interesting, but their interpretation depends on the validation being closed."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick read: this is a solid, useful implementation paper. The new content is the independent CAMB implementation of the Passaglia–Hu EFA, the switch-robustness study, and fresh MCMC constraints with Planck PR4 and DESI BAO. The code is public, the figures are clear, and the comparison with Liu, Hu, and Grin's concurrent implementation is a good-faith cross-check. That part deserves credit.\n\nThe soft spot is exactly what the stress-test flags. The abstract says sub-percent accuracy against exact Klein-Gordon solutions, but the body's validation (Fig. 7, Section III.C) uses the PH EFA at a late switch (m/H*=300, accuracy=3) as the ground truth. That is the same approximation family, so errors common to the EFA—equation of state, sound speed, matching—are invisible to the test. The paper does not reproduce the original Passaglia–Hu calibration against exact KG, and the note-added comparison with Ref. [30] only checks two PH EFA codes against each other. So the headline accuracy claim is not directly demonstrated here. That's a real gap, but it is addressable: a direct exact-KG comparison for the most sensitive masses near 10^-27–10^-28 eV would close it.\n\nThe MCMC limits are new, but the paper itself acknowledges that part of the improvement over Ref. [15] comes from newer data (PR4 and DESI). So the 'improved constraints' are not purely a method effect; the authors are honest about that. The switch-time robustness result (Fig. 2, 0.2% vs 3.9% variation) is a genuine and well-demonstrated advantage of the PH EFA.\n\nOverall: the central implementation is likely correct and the code is a useful community resource. The accuracy claim needs tightening or a direct test before publication. The paper deserves a serious referee; the fix is straightforward.","headline":"Useful CAMB implementation of the PH EFA with honest new constraints, but the headline accuracy claim rests on a self-referential validation.","tokens_in":13368,"tokens_out":1727,"would_cite":true,"duration_ms":14273,"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":"A fast CAMB implementation of the Passaglia–Hu effective fluid approximation claims sub-percent CMB accuracy for ultralight axions and reports tighter dark-matter limits.","keywords":["ultralight axions","effective fluid approximation","Passaglia-Hu approximation","CMB power spectrum","Klein-Gordon equation","cosmological parameter constraints","axion dark matter","AxiCAMB"],"falsifier":"Compute the CMB TT spectrum with the production settings ($m/H_* = 50$, accuracy=1) and compare it with a direct numerical integration of the perturbed Klein–Gordon equation at $m = 5 \\times 10^{-27}$ eV; if the relative difference at $\\ell \\sim 2000$ exceeds roughly one percent, the sub-percent claim is refuted.","tokens_in":12359,"feed_emoji":"🌌","tokens_out":12682,"duration_ms":100831,"temperature":0.7,"pith_summary":"The paper aims to establish that ultralight axions with masses around $10^{-28}$ to $10^{-24}$ eV can be evolved accurately and quickly by replacing the rapidly oscillating scalar field with a calibrated effective fluid. It presents an implementation of the Passaglia–Hu effective fluid approximation inside the Boltzmann code CAMB, in which two slowly varying envelope fields carry the axion dynamics instead of the full Klein–Gordon oscillation. The central claim is that this method matches the CMB temperature power spectrum of the exact Klein–Gordon equation to sub-percent accuracy across that mass range, while allowing an early and cheap switch from field to fluid evolution. Using Planck PR4 CMB data and DESI BAO data, the paper reports $2\\sigma$ upper limits $f_{\\rm ax} < 0.0082$ and $\\Omega_{\\rm ax} h^2 < 0.0010$ at $m = 10^{-28}$ eV, about 50% tighter than earlier standard-EFA analyses. If correct, this removes a known systematic from axion constraints at precisely the masses where current and near-future surveys are most sensitive.","feed_headline":"Axion solver matches exact theory to sub-percent in CMB spectra","feed_subtitle":"Faster, switch-robust axion code tightens dark-matter limits at 10^-28 eV to below 0.0082.","key_machinery":"The load-bearing object is the Passaglia–Hu decomposition of the axion field into two slowly varying auxiliary fields, $\\phi(\\tau) = \\phi_c(\\tau)\\cos(\\tau-\\tau_*) + \\phi_s(\\tau)\\sin(\\tau-\\tau_*)$ with $\\tau = mt$, which filters the fast oscillations out of the equations of motion. From these envelope fields the paper constructs an effective fluid energy density, pressure, and heat flux, and closes the system with a calibrated equation of state $w_{\\rm ax} \\simeq \\frac{3}{2}(m/H)^{-2}$ and a sound speed that carries a $(m/H)^{-2}$ correction. Matching conditions at the switch time suppress spurious oscillatory modes in both the background and perturbations. This machinery is what lets the switch from Klein–Gordon to fluid evolution happen at $m/H_* \\approx 50$ with far less sensitivity to the switch parameter than the standard effective fluid approximation.","core_discovery":"The paper's contribution is a concrete numerical claim: the Passaglia–Hu effective fluid approximation, coded into CAMB as AxiCAMB, reproduces the observable cosmological effects of ultralight axions to sub-percent accuracy in the CMB temperature power spectrum for $m \\in [10^{-28}, 10^{-24}]$ eV. The method decomposes the axion field into amplitude and phase envelope fields whose equations of motion are free of the fast oscillations, then maps them onto an effective fluid with an equation of state and sound speed calibrated to exact Klein–Gordon solutions. The validation in the body compares against a late-switch, high-accuracy version of the same scheme as a ground-truth reference, and finds that the production setting $m/H_* \\approx 50$ with default integrator accuracy stays close to that reference across the mass range, whereas the standard effective fluid approximation requires much later switches. The resulting MCMC analysis yields $2\\sigma$ upper limits $f_{\\rm ax} < 0.0082$ and $\\Omega_{\\rm ax} h^2 < 0.0010$ at $m = 10^{-28}$ eV, with comparable limits at higher masses improving on earlier standard-EFA results.","pith_inferences":["The same auxiliary-field construction is the natural starting point for the full cosine potential and self-interacting axion-like fields; an extension there would let 'extreme axion' scenarios be sampled with the same speed.","Because the central accuracy claim is checked against a late-switch version of the same approximation rather than an independent exact solver, a direct comparison with a true Klein–Gordon integrator would independently test the sub-percent statement.","If the switch-time robustness extends above $10^{-24}$ eV, the method could sharpen forecasts for CMB-S4 and intensity mapping experiments, whose axion sensitivity moves to smaller scales.","The tighter limits combine a more accurate fluid model with newer datasets, so a fixed-data comparison between standard and PH EFA would separate the method's contribution from the data's contribution."],"forward_implications":["Axion constraints in the mass range $10^{-28}$ to $10^{-24}$ eV can be computed without the switch-time systematic that affects the standard effective fluid approximation.","Production MCMC runs can use an early switch with default integrator settings, making percent-level axion limits computationally affordable.","For Planck PR4 plus DESI BAO data, the method gives $f_{\\rm ax} < 0.0082$ and $\\Omega_{\\rm ax} h^2 < 0.0010$ at $m = 10^{-28}$ eV.","The implementation is compatible with the current Boltzmann-code base, so it can be combined with new CMB and large-scale structure likelihoods as they appear."],"supporting_citations":[{"why":"Supplies the auxiliary-field effective fluid approximation whose equations this implementation solves.","marker":"[23]"},{"why":"Defines the standard effective fluid approximation and the AxionCAMB baseline that the paper improves.","marker":"[12]"},{"why":"Provides the standard EFA perturbation treatment and the earlier constraint framework used for comparison.","marker":"[19]"},{"why":"Documents the systematic errors in the standard EFA that motivate the improved approximation.","marker":"[17]"},{"why":"Gives the earlier standard-EFA constraints against which the new MCMC limits are compared.","marker":"[15]"},{"why":"Supplies the Planck PR4 CamSpec likelihood used in the MCMC analysis.","marker":"[13]"},{"why":"Supplies the DESI BAO measurements used in the MCMC analysis.","marker":"[14]"},{"why":"An independent implementation of the same approximation used to cross-check the CMB power spectrum at large switch values.","marker":"[30]"},{"why":"The Boltzmann solver code base into which the new approximation is integrated.","marker":"[24]"}],"fun_headline_variants":["AxiCAMB: sub-percent axion CMB solver tightens bounds","Fast axion fluids match exact theory to <1% in CMB","New axion code speeds CAMB, sharpens dark matter limits","Axion approximation hits sub-percent CMB accuracy","Ultralight axion solver accelerates cosmology with precision"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The sub-percent accuracy claim rests on treating a late-switch run of the same method at $m/H_* = 300$ with accuracy=3 as ground truth, because the paper does not compare directly against exact Klein–Gordon solutions in the body.","fun_headline_variants_meta":{"raw":{"variants":["AxiCAMB: sub-percent axion CMB solver tightens bounds","Fast axion fluids match exact theory to <1% in CMB","New axion code speeds CAMB, sharpens dark matter limits","Axion approximation hits sub-percent CMB accuracy","Ultralight axion solver accelerates cosmology with precision"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000555,"raw_usage":{"total_tokens":2693,"prompt_tokens":1048,"completion_tokens":1645,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":664,"completion_tokens_details":{"reasoning_tokens":1556}},"tokens_in":664,"tokens_out":1645,"duration_ms":11778,"temperature":1.0,"reasoning_tokens":1556,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T15:42:18.669602+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the CMB TT spectrum with the production settings ($m/H_* = 50$, accuracy=1) and compare it with a direct numerical integration of the perturbed Klein–Gordon equation at $m = 5 \\times 10^{-27}$ eV; if the relative difference at $\\ell \\sim 2000$ exceeds roughly one percent, the sub-percent claim is refuted.","supporting_citations":[{"cited_title":"How sound are our ultra-light axion approximations?","cited_arxiv_id":"1909.11094","evidence_quote":"Documents the systematic errors in the standard EFA that motivate the improved approximation."}],"review_version":1}