{"id":"03755603-71af-4f81-851c-763faa026c7d","arxiv_id":"2607.18221","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"A dual-alkali single-cell correlated measurement cancels common-mode magnetic background at high frequencies, improving magnetometer SNR and enabling single-cell model differentiation of exotic spin couplings.","lead":"This paper proposes a correlated measurement of two alkali species in a single vapor cell to suppress common-mode magnetic noise at high frequencies, beyond where conventional comagnetometers work. In a simulated dark-matter search, it shows a ~30-fold background suppression and an order-of-magnitude SNR gain, plus a calibration-free readout of the signal's proton-to-neutron coupling ratio.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The cancellation's central assumption — stability of the magnetic-transfer ratio c=T_B^K/T_B^Rb to ≲3% between Wiener re-estimations — rests on an unvalidated Bloch-model estimate; if c drifts more, the claimed ×28 suppression and ×15 SNR gain fail.","rationale":"The reader's weakest assumption — stability of the transfer ratio c to ≲3% and the quantitative accuracy of the coupled Bloch equations — is exactly the condition on which the central cancellation claim rests. The paper is internally consistent and the Wiener subtraction is self-calibrating with respect to absolute calibration errors, which is a genuine strength. However, the stability requirement is stated explicitly in the Supplemental Material and is supported only by a single Bloch-model derivative with no error bar. A small drift in c is not a speculative failure mode; it is the natural consequence of temperature, density, and pump drifts. The proposed experimental check directly measures this drift and would settle whether the cancellation reaches the claimed floor. Since this is a validation gap rather than an identified inconsistency, the appropriate verdict remains CONDITIONAL as the reader assigned. I agree with the reader that no red flags invalidate the central idea, but the missing experimental demonstration of c stability is the key unproven step.","tokens_in":14523,"tokens_out":5967,"duration_ms":52254,"concrete_test":"Experimental test: In a 87Rb–39K–3He comagnetometer operating under the conditions of Table I (T≈180 °C, densities, pump rates, α=0.578), apply a calibrated oscillating magnetic field at a frequency in the 70–120 Hz band (e.g. 100 Hz) and measure the complex response ratio c = T_B^K/T_B^Rb over repeated 200 s windows for at least one hour, while actively stabilizing the compensation field and temperature as described in Ref. [30]. If the fractional drift in c between successive windows exceeds 3%, the claimed ×28 suppression and ×15 SNR gain are not achievable in practice, and the central claim would need to be weakened or conditioned on a faster re-estimation scheme.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim is that a Wiener-filtered subtraction of two same-cell alkali channels cancels the correlated magnetic background down to the uncorrelated readout floor, giving a factor-28 noise reduction and ×15 SNR gain (Fig. 3). The cancellation's residual from the correlated background is set by the fractional drift δc in the transfer ratio c ≡ T_B^K/T_B^Rb between successive re-estimations of the Wiener filter H. The paper itself states (Supplemental Material §6, 'Response-ratio stability') that reaching the 0.22 fT/√Hz floor requires δc ≲ 3% over the ~200 s record length. The only support for this stability is a Bloch-model estimate d ln|c|/dT ≈ 5×10^-4 K^-1 at 100 Hz, with no quoted uncertainty and no experimental validation. If the real cell exhibits larger c drift — e.g. due to temperature gradients, density or pump variations, or imperfect compensation-field feedback — the correlated background leaks into the residual and the suppression factor degrades proportionally. Because this stability condition is the load-bearing requirement for the headline sensitivity, it must be demonstrated in a real 87Rb–39K–3He cell before the central claim is accepted. The Bloch-model accuracy enters as the foundation of this estimate, making the stability claim doubly unverified.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a dual-alkali (87Rb, 39K) comagnetometer in a single cell containing 3He, in which two alkali species are read out simultaneously and combined by a Wiener-filtered subtraction. The authors claim that this correlated measurement cancels the correlated magnetic background at high frequencies, where the usual noble-gas self-compensation fails, recovering the uncorrelated readout floor. They report a simulated ~28–30-fold background suppression and a ×15 SNR gain for a dark-matter transient, and a calibration-free inter-species phase difference Δφ that carries the neutron-to-proton coupling ratio R and enables model differentiation. The theory is based on coupled Bloch equations with literature parameters, and the results are presented as a theoretical demonstration with a detailed noise budget.","tokens_in":14880,"tokens_out":19720,"duration_ms":167332,"significance":"If the claims hold, this is a valuable and original contribution: it offers a path to extend comagnetometric common-mode rejection to frequencies above the noble-gas compensation point, improves the sensitivity of spin-based exotic-field searches, and, for the first time, proposes extracting the nucleon coupling ratio from a single vapor cell without inter-device comparison. The use of an experimentally measured Wiener filter rather than a model-dependent calibration is elegant, and the paper is transparent about its assumptions. The main ideas are falsifiable and the simulation setup is described in enough detail to be reproduced. The principal caveat is that the quantitative predictions rest on a Bloch model and on an assumed stability of the magnetic transfer-ratio that is not yet experimentally validated.","major_comments":[{"comment":"The compensation field formula has inconsistent indices: the terms −λ_Rb,He M_Rb^0 P_Rb^z − λ_K,He M_K^0 P_K^z are alkali-induced fields on 3He, not fields on the alkalis, and no alkali–alkali Fermi-contact terms appear. The accompanying text says only the 3He-on-alkali term is interpolated, which conflicts with the equation. Because the operating point α and all subsequent results (antiresonance, phase map) depend on this formula, please correct it or clarify the intended compensation condition.","section":"Eq. (6) and Supplemental §2"},{"comment":"The central claim that the correlated background is cancelled to the 0.22 fT/√Hz readout floor requires δc ≲ 3% stability of c = T_B^K/T_B^Rb over the Wiener re-estimation interval. The only quantitative support is a Bloch-model derivative d ln| c |/dT ≈ 5×10^-4 K^-1 with no quoted uncertainty and no assessment of other drift channels (density, pump rate, gradients, B_z feedback). Because this is the load-bearing assumption for the headline suppression, provide a sensitivity analysis over plausible parameter drifts or explicitly state the result as conditional on this stability.","section":"Supplemental §6, 'Response-ratio stability'"},{"comment":"The quantitative predictions (×28 suppression, ×15 SNR gain) are obtained from a single set of cell parameters (Table I) and a single noise model, without error bars. The retention g(f) and suppression S(f) depend on the Bloch-model parameters (λ, κ, densities), whose literature values carry few-percent uncertainties. Please propagate these uncertainties or at least demonstrate robustness of the gain to parameter variations, since the quantitative claims are a central part of the paper.","section":"Fig. 3 and main text §5"}],"minor_comments":[{"comment":"The abstract says 'thirtyfold background suppression' while Fig. 3(c) shows a factor of 28; please make the numbers consistent.","section":"Abstract vs Fig. 3(c)"},{"comment":"The two-channel readout floor is quoted as 0.22 fT/√Hz, but uncorrelated 0.17 fT/√Hz per-channel noise added in quadrature gives √2×0.17 ≈ 0.24 fT/√Hz. Clarify the apparent discrepancy.","section":"Supplemental §6, readout floor"},{"comment":"The antiresonance ranges are said to be 'tabulated in Ref. [29]'; please refer explicitly to Table III to avoid ambiguity.","section":"Supplemental §4"},{"comment":"The synthetic-field calibration is referenced to Ref. [24] but is not summarized. Please add a sentence describing how the synthetic exotic field's coupling ratio R is set and varied, since this underpins the calibration-free claim.","section":"Main text §4 and Supplemental §4"}],"recommendation":"major_revision","confidential_remarks":"The manuscript presents a conceptually sound and useful theoretical proposal, but the headline quantitative result is conditioned on the stability of the magnetic transfer-ratio c, which is not experimentally characterized and is defended by a single Bloch-model estimate without uncertainty. The index problem in Eq. (6) must be fixed. I do not see grounds for rejection; the idea is publishable after a careful revision that addresses the stability analysis and the equation typo."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First thing you should know: this is a clever proposal, not a demonstrated result. The genuinely new idea is correlating two same-cell alkali species (87Rb and 39K) and subtracting their outputs with a measured Wiener transfer function to cancel correlated magnetic noise at high frequencies, where noble-gas self-compensation stops working. That is a real combination I have not seen before. The inter-species phase difference Δφ as a calibration-free readout of the neutron-to-proton coupling ratio R is also new and well-motivated, and the paper shows how it can distinguish KSVZ from DFSZ-like couplings.\n\nThe paper is internally consistent. The Bloch model is first-principles with literature parameters; nothing is fitted to the headline numbers. The supplemental material goes beyond the usual hand-waving: it budgets light-shift noise, differential gradient sampling, spin-projection noise, and the response-ratio stability that the cancellation relies on. The figures are honest—they quote the retention factor g≈0.53, which means the recovered signal is attenuated, not just amplified in SNR.\n\nThe soft spots are proportionate. There is no experimental validation; this is a proposal, and the quantitative predictions (×28 suppression, ×15 SNR gain) rest on one set of cell parameters and a specific noise model. The load-bearing stability assumption—that the magnetic transfer ratio c is stable to better than ~3% over the timescale between Wiener re-estimations—is supported only by an unvalidated Bloch-model estimate with no quoted uncertainty. The paper argues persuasively that the requirement is loose, but it is not a measurement. If c drifts more in a real cell, the suppression degrades proportionally. That is not a fatal flaw in the idea, but it means the headline numbers should be read as simulated under assumed stability. The absence of code and data also limits independent reproduction.\n\nWho this is for: people in exotic-spin dark matter searches, comagnetometer development, and precision metrology. It deserves a serious referee. I would send it out, and ask the authors to release the simulation code or a full parameter dump, and to address how the transfer-ratio stability would be verified in practice.","headline":"A genuinely new dual-alkali correlated readout that plausibly cancels high-frequency magnetic noise; the central stability assumption is untested but the proposal is honest and deserves refereeing.","tokens_in":15353,"tokens_out":3927,"would_cite":true,"duration_ms":33796,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Two alkali species in one cell cancel magnetic noise at high frequencies and read out the signal's coupling structure.","keywords":["comagnetometry","magnetic noise cancellation","dual-alkali cell","spin-exchange optical pumping","Wiener filter","axion dark matter","coupling ratio","precision measurement"],"falsifier":"Run the proposed dual-alkali cell with a calibrated 70–120 Hz magnetic tone and measure the residual after Wiener subtraction over a 200 s record; if the residual floor stays above the combined readout and spin-projection noise of about 0.22 fT/√Hz because the transfer-ratio drifts by more than about 3%, the central cancellation claim fails.","tokens_in":14440,"feed_emoji":"🧲","tokens_out":6719,"duration_ms":63016,"temperature":0.7,"pith_summary":"Comagnetometers suppress magnetic noise by self-compensation with a noble-gas spin species, but this only works at low frequencies. This paper proposes a way to cancel the magnetic background at higher frequencies too: read out two alkali species in the same cell and subtract one channel from the other using a measured least-squares transfer function. Because both alkalis see the same magnetic field, the magnetic background is common-mode and cancels, while a non-magnetic signal that couples to the two species differently survives. In simulations of a transient axion-like dark-matter signal, the correlated background drops by about a factor of 28 in the 70–120 Hz band, the noise floor reaches the uncorrelated readout level, and the signal-to-noise ratio improves by a factor of 15. The paper also shows that the phase difference between the two alkali responses is calibration free and encodes the neutron-to-proton coupling ratio of the new field, allowing a single cell to distinguish between dark-matter model classes.","feed_headline":"Two alkalis in one cell cut magnetic noise 30-fold","feed_subtitle":"Correlated readout of 87Rb and 39K suppresses common magnetic noise and reveals the signal's neutron–proton coupling","key_machinery":"The carrying mechanism is the shared magnetic transfer function of the two alkali channels, used in two observables. First, the field-referred signals are combined as r = x_K − Ĥ x_Rb, where Ĥ is a least-squares (Wiener) transfer function estimated from signal-free records; any error in the magnetic transfer functions folds into Ĥ and cancels, making the subtraction self-calibrating. Second, the inter-species phase difference Δφ is formed from the two Faraday-rotation phases; multiplicative prefactors cancel in each phase, making Δφ calibration free and independent of common-mode intensity fluctuations. The response of the coupled 87Rb–39K–3He spin system is modelled by coupled Bloch equatio","core_discovery":"The central claim is that a correlated measurement of two same-cell alkali species makes a differential readout an excellent common-mode rejection filter for the magnetic background in principle at all frequencies within the linear regime. Because 87Rb and 39K share the same magnetic field and electronic gyromagnetic ratio, a magnetic perturbation enters both channels with nearly identical transfer functions, while an exotic spin-dependent field enters through species-specific effective couplings that depend on the nuclear spin content. Subtracting the 87Rb-referred channel from the 39K-referred channel with an adaptively measured Wiener coefficient cancels the correlated magnetic background","pith_inferences":["One extension the paper does not develop: the Wiener residual itself could serve as a live monitor of transfer-ratio drift, triggering recalibration when the residual floor rises, which would relax the stability requirement in practice.","Beyond the paper: the claimed suppression factor scales with the ratio of correlated to uncorrelated noise, so in a less shielded or noisier environment the gain could be larger; this might enable useful high-frequency sensitivity with lighter shielding.","As an editorial test of the model, the phase map Δφ(ω;R) could be measured with a synthetic field at several operating points and the recovered R compared across them; consistency would support the Bloch-model assumptions, and inconsistency would reveal where the model needs correction."],"forward_implications":["If the cancellation works as simulated, any magnetometer search limited by correlated magnetic noise above the noble-gas Larmor frequency becomes readout-limited instead, without needing to improve the magnetic shields.","A detected exotic signal would carry a calibration-free measurement of the neutron-to-proton coupling ratio R from a single vapor cell; comparing this value to model predictions can distinguish between the main axion model classes or rule both out.","The method is not specific to axions: any perturbation that couples to the two alkali species differently than a magnetic field does will survive the common-mode subtraction, including electron electric dipole moments and species-specific light shifts.","The tunable antiresonance, with its exact-null locus, provides a second experimental handle: changing the applied longitudinal field moves the null frequency, which can verify the coupling model and enhance sensitivity near the null.","Because the subtraction requires only modest additions to a standard comagnetometer—a second probe laser, its polarimetry chain, and a dichroic beamsplitter—existing devices could be retrofitted."],"fun_headline_variants":["Correlated alkali pair cancels magnetic noise 30-fold","Two-species comagnetometer suppresses background 30x","Calibration-free magnetic noise cancellation via alkali duo","Same-cell alkali pair makes magnetic noise vanish at all frequencies"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The cancellation only works if the ratio of the two alkali channels' magnetic transfer functions stays constant within a few percent over the time between recalibrations; the paper argues spin-exchange locking provides this stability, but it is not yet demonstrated in experiment.","fun_headline_variants_meta":{"raw":{"variants":["Correlated alkali pair cancels magnetic noise 30-fold","Two-species comagnetometer suppresses background 30x","Calibration-free magnetic noise cancellation via alkali duo","Same-cell alkali pair makes magnetic noise vanish at all frequencies"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000957,"raw_usage":{"total_tokens":3864,"prompt_tokens":641,"completion_tokens":3223,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":385,"completion_tokens_details":{"reasoning_tokens":3158}},"tokens_in":385,"tokens_out":3223,"duration_ms":22665,"temperature":1.0,"reasoning_tokens":3158,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T15:37:11.660908+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the proposed dual-alkali cell with a calibrated 70–120 Hz magnetic tone and measure the residual after Wiener subtraction over a 200 s record; if the residual floor stays above the combined readout and spin-projection noise of about 0.22 fT/√Hz because the transfer-ratio drifts by more than about 3%, the central cancellation claim fails.","supporting_citations":[],"review_version":1}