{"id":"d0f5b277-0ecf-4b84-9abb-8bb526822a48","arxiv_id":"2506.23983","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"LEKID arrays with a 45 degree polarizer achieve 0.1 degree (6.5 arcmin) statistical precision in polarization angle reconstruction at 150 GHz in the lab.","lead":"Two arrays of superconducting LEKID detectors, split by a 45 degree polarizer, reconstruct the polarization angle of 150 GHz light with about 6.5 arcminutes of statistical uncertainty in a laboratory test bench. This is a step toward showing that this detector technology can meet the strict polarization calibration requirements of next-generation cosmic microwave background experiments searching for signs of inflation.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The quoted 6.5 arcmin is the statistical fit uncertainty, not the absolute polarization-angle accuracy; with P2 known only to 0°±1° (Sect. 4.3), the absolute floor is about 1°, so the abstract/conclusion claim of meeting next-generation CMB requirements is not supported by the reported data.","rationale":"The reader's weakest assumption is exactly right. The central claim is that LEKIDs in this configuration achieve the absolute polarization-angle precision required for next-generation CMB B-mode measurements. What the experiment actually constrains is the relative response of two arrays as a function of source-polarizer rotation, fitted with Eq. 11 to yield β. The fit is insensitive to a common unknown rotation of the whole polarimetric reference frame; P1's 0.1° microscope calibration and P2's 1° leveling define that frame. A 1° unknown is not covered by the 0.1° statistical error, and it changes the conclusions: at 1°, E-to-B leakage in Eq. 13 would bias r at the ~0.01 level, similar to Planck, not at 0.001. The paper's own Sect. 5.1 and Conclusions acknowledge that alignment/calibration dominates the uncertainty, but the abstract and conclusions still present 6.5 arcmin as a standalone result. This is an overclaim in framing rather than a flaw in the detector measurement. The reader already issued CONDITIONAL, and this stress-test reinforces that condition: the paper should report a total error budget including the 1° reference floor, and state explicitly that the 0.1° result is a repeatability/precision result pending absolute calibration. No evidence of internal inconsistency or fabrication; the raw measurement and NEP characterization are credible and reproducible in principle.","tokens_in":15925,"tokens_out":4732,"duration_ms":50923,"concrete_test":"Compute the full quadrature error budget from Sect. 4.3: σ_total² = σ_P2_level² (1°) + σ_P1_microscope² (0.1°) + σ_encoder² (0.01°) + σ_fit² (0.1°), giving σ_total ≈ 1.0°. If this exceeds 6.5 arcmin, the headline claim is only repeatability. To settle it experimentally, re-align P2 with an autocollimator/theodolite referenced to the optical axis to ≤0.02°, repeat the α-rotation sequence, and refit β; if the recovered β shifts by more than 0.1° relative to the assumed 0° prior, the absolute-angle claim fails, and if it shifts by less, the 1° uncertainty was conservative.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's headline result, 'polarization angle reconstructed with an uncertainty of 6.5 arcmin' and the conclusion that 0.1° precision meets CMB requirements, rests on presenting a fit precision as an absolute calibration. In Sect. 4.3, the source polarizer P1 is aligned by microscope to 0.1°, while the cold polarizer P2 is only 'empirically leveled' to 0°±1°. The Eq. 11 fit in Sect. 5.1 gives β=0.8°±0.1°, and the paper explicitly says the ±0.1° 'accounts for the reproducibility and intrinsic dispersion' among KIDs, not for the 1° prior on P2. That 1° term is a systematic floor on any absolute polarization angle reported by the setup, not a scatter that averages down. Eq. 13 and Fig. 12 use Δψ as an absolute miscalibration; plugging a 1° floor into Eq. 15 raises the r bias to roughly Planck-level (r~0.01), not the quoted r~0.001. The conclusions themselves admit 'the current uncertainty in polarization angle reconstruction is primarily influenced by the alignment and calibration accuracy of the experimental setup,' so the abstract and conclusions overstate what has been demonstrated.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript presents a laboratory characterization of two filled-array LEKID cameras operating at 150 GHz, configured with a 45° linear polarizer between them. Using a sky simulator with a rotating wire-grid polarizer (P1) as an external reference, the authors fit a Mueller-matrix model to aperture-photometry measurements to determine the orientation of the cold polarizer (P2). They report a reconstructed polarization-angle uncertainty of 6.5 arcmin (0.1°), a near-photon-noise NEP, and a forecast that this precision would bias the tensor-to-scalar ratio at r ~ 0.001. The conclusions state that the setup meets the polarization-angle requirements of next-generation CMB experiments.","tokens_in":16155,"tokens_out":10093,"duration_ms":104347,"significance":"If the 0.1° value is a true total absolute uncertainty, this would be a meaningful step for LEKID-based CMB polarimetry, especially compared with the ~1° statistical angle precision reported for NIKA2. The paper is also valuable for its detailed description of the test bench, including the sky simulator, Martin-Puplett interferometer, aperture photometry, MCMC fitting, and explicit discussion of cross-polarization and calibration limitations. However, the headline precision is not placed in a complete uncertainty budget, so the central CMB-readiness claim is not yet fully established.","major_comments":[{"comment":"The manuscript reports that the polarization angle was reconstructed with an uncertainty of 6.5 arcmin and states in the Conclusions that a precision of 0.1° meets next-generation CMB requirements. However, §5.1 explicitly defines the ±0.1° as the statistical uncertainty accounting for reproducibility and intrinsic dispersion among the KIDs, while §4.3 gives only the pre-calibration accuracy: P1 is aligned by microscope to 0.1°, and P2 is empirically leveled to 0°±1°. Although the fit may be used to calibrate P2 relative to P1, the paper never propagates the P1 reference uncertainty and the P2 placement prior into a total absolute uncertainty for the reconstructed angle. As written, the abstract and Conclusions conflate fit precision with absolute calibration accuracy; the Conclusions' own statement that the current uncertainty is dominated by the alignment and calibration of the setup undercuts the headline claim. Please provide an explicit error budget separating statistical, P1-reference, P2-alignment, and cross-polarization contributions, and adjust the abstract and conclusions accordingly.","section":"Abstract, §4.3, §5.1, Conclusions"},{"comment":"The forecast for the tensor-to-scalar ratio bias uses Δψ = 0.1° as the absolute polarization-angle error. Since the paper does not establish that 0.1° is the total absolute uncertainty, the 'this work' point in Fig. 12 is not conservative. For example, adding the 0.1° P1 reference uncertainty in quadrature with the 0.1° statistical term gives Δψ ≈ 0.14°, which via Eq. (16) approximately doubles the E-to-B leakage relative to Δψ = 0.1°, and a 1° systematic floor would bring the bias to roughly the Planck level (r ~ 0.01). The forecast should be recomputed using a properly propagated total uncertainty and should include a sensitivity curve r(Δψ) over the range allowed by the setup.","section":"§5.3, Eqs. (13)-(16), Fig. 12"}],"minor_comments":[{"comment":"The reference temperatures provided by thermometers are treated as exact values without associated uncertainties; please state the estimated thermometer uncertainty and its effect on the derived effective sky-simulator temperature.","section":"§3.1"},{"comment":"The text 'can also be expresses as' should read 'can also be expressed as'.","section":"Eq. (4)"},{"comment":"The x-axis label appears to be missing the variable Δψ; please add it and specify the units explicitly.","section":"Fig. 12"},{"comment":"Section 2.2.1 states a temperature stability of 0.1 mK at the 100 mK stage, while the Conclusions mention fluctuations within 1 mK during a typical measurement session; please reconcile these values or specify the different timescales.","section":"§2.2.1 and Conclusions"},{"comment":"The parasitic cross-polarization component k(T) is estimated as approximately 1% without an associated uncertainty; please provide an uncertainty or state explicitly that this is an order-of-magnitude estimate.","section":"§5.2"}],"recommendation":"major_revision","confidential_remarks":"The paper's own conclusions admit that the current uncertainty is dominated by setup calibration, which makes the abstract's unqualified 6.5 arcmin claim problematic. The editors may wish to ask the authors for a revised abstract and a full uncertainty budget before considering publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: the paper is a solid lab characterization of LEKID polarization angle reconstruction, and the 6.5 arcmin statistical uncertainty is credible. The abstract and conclusions, however, present that as meeting next-generation CMB requirements, and that only holds if you ignore the 1° uncertainty on the cold polarizer's absolute orientation. That is a real overclaim.\n\nWhat's new: a dedicated testbed with a sky simulator, Martin-Puplett interferometer, and two perpendicular LEKID arrays behind a 45° polarizer. The fit of β = 0.8° ± 0.1° from the data is clean, and the treatment of the parasitic cross-polarization term is sensible. The NEP is close to photon noise within the quoted uncertainties. As a demonstration of how well the angle can be reconstructed from the detector response in a controlled setup, it is a useful step forward from NIKA2's ~1° on-sky result.\n\nThe soft spot is exactly where the stress-test lands. The 0.1° is the scatter of the fit across the selected pixels, not the absolute calibration of the instrument. P2 is only leveled to 0° ± 1°, and the conclusions say the current uncertainty is dominated by alignment and calibration. So the floor on absolute angle accuracy is about 1°, not 6.5 arcmin. Their own Eq. 15 and Fig. 12 show that a 1° error gives r ~ 0.01, not the r ~ 0.001 they put forward. The fix is straightforward: report the total error budget, state clearly that this is a precision/repeatability result, and separate that from the absolute calibration requirement.\n\nThe pixel selection (30% best) is a reasonable choice for a method validation, but it should be more prominent so no one reads it as full-array performance. The NEP comparison is loose but not wrong.\n\nFor a reader: this is for instrumentation people who care about KID arrays and CMB polarimetry calibration. It deserves a serious referee, but the authors should not get away with the current abstract. I would send it to review with a major-revision recommendation.","headline":"A real precision measurement in the lab, but the CMB-ready claim only works if you treat statistical scatter as absolute calibration, which it is not.","tokens_in":16788,"tokens_out":2607,"would_cite":true,"duration_ms":29110,"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":"Two LEKID arrays measure polarization angle to 6.5 arcmin","keywords":["kinetic inductance detectors","LEKID","CMB polarization","polarization angle calibration","E-to-B leakage","150 GHz","filled-array focal plane","sky simulator"],"falsifier":"Measure the absolute orientation of the cold polarizer P2 independently to better than 0.05 degrees, for instance with a theodolite sighting through the optical path, and repeat the polarization-curve fit; if the independently measured angle differs from the fitted $\\beta = 0.8^\\circ$ by more than the claimed uncertainty, then 6.5 arcmin is a relative precision rather than an absolute angle accuracy.","tokens_in":15700,"feed_emoji":"📡","tokens_out":11299,"duration_ms":109864,"temperature":0.7,"pith_summary":"Lumped Element Kinetic Inductance Detectors (LEKIDs) are superconducting resonators that both absorb radiation and resonate at a frequency, and in a filled array they cover the focal plane directly without feedhorns. The paper tries to establish that two such arrays, mounted perpendicular to each other with a wire-grid polarizer at 45 degrees splitting the beam, can reconstruct the angle of incoming linear polarization precisely enough for next-generation CMB experiments. Using a 100 mK cryostat coupled to a sky simulator and a rotated polarized source at 150 GHz, the authors report a reconstructed polarization-angle uncertainty of 6.5 arcmin, about 0.1 degrees. This matters because a polarization-angle error of order 0.1 degrees is the level at which E-mode power starts leaking into B-modes and biasing the tensor-to-scalar ratio $r$ in CMB polarization searches. If the result holds, LEKID filled arrays become a credible detector technology for the absolute polarization calibration demanded by future B-mode experiments.","feed_headline":"LEKID arrays hit 6.5-arcmin polarization angle","feed_subtitle":"Two detector arrays split by a 45-degree polarizer meet the 0.1-degree target for CMB B-mode searches.","key_machinery":"The load-bearing mechanism is the two-arm polarization splitter formed by the two mutually perpendicular LEKID arrays and the 45-degree wire-grid polarizer P2 between them: the transmission array records one linear polarization while the reflection array records the orthogonal one, giving a 90-degree phase separation. The quantitative core is the Mueller-matrix model $V = 1 + \\sin 2\\alpha \\cos 2\\beta + \\cos 2\\alpha \\sin 2\\beta$, where $\\alpha$ is the rotated source polarizer's angle and $\\beta$ is the cold polarizer's orientation; fitting this model to the measured intensity-versus-angle curves with a Monte Carlo Markov Chain returns $\\beta$ and its uncertainty. This model is what converts raw detector maps into a polarization angle, and the width of its fitted parameter distribution is the claimed 0.1-degree precision.","core_discovery":"The paper's central claim is that a focal plane built from two perpendicular LEKID arrays, with a wire-grid polarizer at 45 degrees splitting the incoming beam, measures the linear polarization angle of a 150 GHz source with an uncertainty of 6.5 arcmin (about 0.1 degrees), meeting the target that next-generation CMB experiments set for absolute polarization calibration. The claim is supported by a laboratory polarization curve: the source polarizer was rotated in 15-degree steps, the detected flux was integrated by aperture photometry on the two arrays, and a Mueller-matrix model fitted to those curves returned the cold polarizer angle $\\beta = 0.8^\\circ \\pm 0.1^\\circ$. The same measurements give a cross-polarization of about 1% and a pixel noise equivalent power of about $1.5 \\times 10^{-16}$ W per root hertz, within roughly a factor 1.1 of the photon-noise limit. The paper concludes that LEKID filled arrays can meet the angle-precision requirement for CMB B-mode searches.","pith_inferences":["Beyond the paper: the 0.1-degree figure is the statistical width of the fitted cold-polarizer angle, and the absolute accuracy inherits the roughly one-degree leveling uncertainty of that polarizer, so the headline precision is best read as relative reproducibility until an externally calibrated absolute reference is used.","Beyond the paper: the same Mueller-matrix procedure should transfer to on-sky operation with a rotating half-wave plate, where the sky itself acts as the rotating source, providing a direct end-to-end test of the angle calibration.","Beyond the paper: because the paper identifies setup alignment as the dominant uncertainty, replacing the microscope-leveled source polarizer with an absolutely calibrated polarizing source is a straightforward improvement path that could push the stated uncertainty below 0.1 degrees and is testable with the existing apparatus."],"forward_implications":["If the central claim holds, LEKID filled arrays become viable focal planes for CMB experiments targeting tensor-to-scalar ratios around $r \\sim 10^{-3}$, where a polarization-angle error of 0.1 degrees keeps E-to-B leakage at the multipoles that matter.","The demonstrated test bench can serve as a qualification facility for other millimeter-wave cameras, since it produces diffraction-limited intensity and polarization maps under realistic ground-based optical loading.","The measured noise equivalent power, close to the photon-noise limit, implies that a ground-based 150 GHz instrument using these arrays would not be limited by detector noise in the 1-10 Hz band used for polarization modulation.","The path to wider deployment is the paper's stated plan: extend the analysis from the best 30% of pixels to full arrays, scale toward roughly 30,000 detectors, and validate the result at 250 GHz."],"supporting_citations":[{"why":"Sets the 0.1-degree polarization-angle target by quantifying E-to-B leakage at multipoles 80 to 100 for $r \\sim 10^{-3}$.","marker":"Rosset et al. 2010"},{"why":"Provides the plus-or-minus-one-degree polarization-angle uncertainty baseline for LEKID polarimetry that the present method must beat.","marker":"Ritacco et al. 2017"},{"why":"Gives the Crab Nebula as a reference source for polarization calibration, anchoring the astrophysical baseline.","marker":"Ritacco et al. 2022"},{"why":"Motivates external absolute calibration by showing that self-calibration relies on assumptions about the CMB polarization signal.","marker":"Keating et al. 2013"},{"why":"Establishes the filled-array LEKID concept that the present camera builds on.","marker":"Monfardini et al. 2010"},{"why":"Supplies the common-mode subtraction and decorrelation procedure used to turn time-ordered detector data into maps.","marker":"Perotto et al. 2020"},{"why":"Provides the Monte Carlo Markov Chain fitting tool that returns the fitted polarizer angle and its uncertainty.","marker":"Foreman-Mackey et al. 2013"}],"fun_headline_variants":["LEKID arrays hit 6.5 arcmin polarization angle","Two LEKID arrays + 45° polarizer reach CMB precision","Polarization angle uncertainty 6.5 arcmin with LEKIDs","LEKID focal plane meets CMB B-mode angle target","6.5-arcmin polarization angle from LEKID detectors"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The result treats the fitted 0.1-degree scatter as the total polarization-angle uncertainty, even though the paper states that the cold polarizer's orientation was only set to within one degree, so if that alignment uncertainty dominates, the absolute angle accuracy is about one degree rather than 0.1 degrees.","fun_headline_variants_meta":{"raw":{"variants":["LEKID arrays hit 6.5 arcmin polarization angle","Two LEKID arrays + 45° polarizer reach CMB precision","Polarization angle uncertainty 6.5 arcmin with LEKIDs","LEKID focal plane meets CMB B-mode angle target","6.5-arcmin polarization angle from LEKID detectors"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000427,"raw_usage":{"total_tokens":2185,"prompt_tokens":946,"completion_tokens":1239,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":562,"completion_tokens_details":{"reasoning_tokens":1145}},"tokens_in":562,"tokens_out":1239,"duration_ms":10522,"temperature":1.0,"reasoning_tokens":1145,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T21:27:29.992408+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the absolute orientation of the cold polarizer P2 independently to better than 0.05 degrees, for instance with a theodolite sighting through the optical path, and repeat the polarization-curve fit; if the independently measured angle differs from the fitted $\\beta = 0.8^\\circ$ by more than the claimed uncertainty, then 6.5 arcmin is a relative precision rather than an absolute angle accuracy.","supporting_citations":[{"cited_title":"2010, A&A, 520, A13","cited_arxiv_id":null,"evidence_quote":"Sets the 0.1-degree polarization-angle target by quantifying E-to-B leakage at multipoles 80 to 100 for $r \\sim 10^{-3}$."},{"cited_title":"2017, A&A, 599, A34","cited_arxiv_id":null,"evidence_quote":"Provides the plus-or-minus-one-degree polarization-angle uncertainty baseline for LEKID polarimetry that the present method must beat."},{"cited_title":"2022, in European Physical Journal Web of","cited_arxiv_id":null,"evidence_quote":"Gives the Crab Nebula as a reference source for polarization calibration, anchoring the astrophysical baseline."},{"cited_title":"G., Shimon, M., & Yadav, A","cited_arxiv_id":null,"evidence_quote":"Motivates external absolute calibration by showing that self-calibration relies on assumptions about the CMB polarization signal."},{"cited_title":"F., et al","cited_arxiv_id":null,"evidence_quote":"Supplies the common-mode subtraction and decorrelation procedure used to turn time-ordered detector data into maps."}],"review_version":1}