{"id":"3136bcab-9db5-4a9c-89c7-ba39a633af8b","arxiv_id":"2511.08797","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Reversing a quadrupole trap's polarity makes the cloud's displacement a differential magnetic-field signal, giving milli-Gauss-level two-axis sensing from absorption images alone.","lead":"This paper shows that a cold atom cloud in a magnetic trap shifts when an outside magnetic field is present, and that flipping the trap's magnetic polarity and comparing images reveals the field's direction and strength. The technique needs only standard absorption imaging, so it could help labs cancel stray magnetic fields without adding spectroscopy hardware.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified: the gravity-induced centroid offset cancels by translation invariance, and weak gradients enter only at second order.","rationale":"The reader identified the gravity-skewness offset as the weakest assumption, but our analysis shows that the offset cancels exactly under the stated conditions because the potential in the frame comoving with the magnetic zero is translationally invariant. The paper's Section 2.4 and Fig. 5 provide both theoretical and experimental support for this cancellation. The inhomogeneity concern from Section 3.3 is also weaker than it first appears: to first order in the external gradient, the differential signal is unaffected. The genuine limitations—two-dimensional sensitivity, lack of an absolute accuracy test, and unreleased data—are appropriately reflected in the reader's CONDITIONAL verdict, but they do not rise to a load-bearing flaw in the central claim. Thus no change to the verdict is warranted; the work remains a conditional acceptance pending the promised data and calibration.","tokens_in":11608,"tokens_out":18746,"duration_ms":194475,"concrete_test":"Apply a known homogeneous field via the bias coils (with coil constants α_i determined from the measured current-to-field relationship), record differential displacements for both quadrupole polarities over a range of applied fields (e.g., ±100 mG in steps), and compare the inferred field from Eq. (3) against the applied values. A slope error greater than the claimed 5 mG resolution would indicate a systematic bias in the centroid cancellation or the quadrupole model.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The reader's weakest assumption is that the 2D Gaussian centroid differs from the magnetic-field zero by a polarity-independent offset. This assumption is actually well-founded. Shifting to u = r − r0, the total potential is U(u) = μ|Q M u| + m g u_z, which is independent of B_ext and of the sign of Q (since |B| is unchanged by polarity reversal). Therefore the density profile—and any fit bias due to gravity skewness—is translated rigidly with r0, so the centroid offset is a constant vector that cancels in the difference. Figure 5 directly supports this: at the inferred compensation currents (B_ext = 0), the two polarity ensembles overlap, indicating equal offsets. Similarly, a constant external gradient G affects the differential displacement only at second order in G/Q (the linear terms cancel), so the abstract's claim of insensitivity to weak inhomogeneities is valid to first order. The remaining concerns are quantitative rather than conceptual: no absolute calibration of the field scale has been demonstrated, and the raw images are not yet public. These do not undermine the central buoy mechanism.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces a differential magnetic sensing method for cold-atom quadrupole traps. For an ideal quadrupole field, a homogeneous external field displaces the magnetic-field zero by r0 = [ẑ∘ẑ − 2(x̂∘x̂+ŷ∘ŷ)] B_ext/(2Q), and reversing the quadrupole polarity Q flips this displacement. Measuring the difference of the trapped-cloud positions under opposite polarities therefore gives a linear, directional estimate of the transverse external field. The authors demonstrate the effect in a ⁸⁷Rb apparatus by varying compensation currents, fitting the polarity-resolved cloud centers, identifying zero-crossing compensation currents, and estimating a ~5 mG field uncertainty from the Allan deviation of the center determination. They also discuss gravity-induced centroid offsets, effects of magnetic-field inhomogeneities, and a possible extension to 3D sensing.","tokens_in":11873,"tokens_out":11685,"duration_ms":125994,"significance":"If the claims hold, this is a genuinely simple addition to standard cold-atom infrastructure: it yields vector information about stray fields using only absorption imaging and current polarity reversal, without spectroscopy or internal-state coherence. The central formula Eq. (3) is a correct, parameter-free geometric derivation, which is a clear strength. The experimental data in Figs. 2–5 show the expected linear and directional response, and Fig. 4 provides an out-of-sample consistency check through the 'stopping' condition in Fig. 5. The authors also give a transparent error analysis based on Allan deviation and support their approximations with numerical field calculations. The gravity-cancellation concern raised in the review is actually resolved by translation invariance: in shifted coordinates u = r − r0, the potential is μ|Q M u| + m g u_z, independent of B_ext and of the sign of Q, so the centroid offset is a constant vector and cancels in the difference. The main weaknesses are the misleading treatment of inhomogeneities in Sec. 3.3, an explicit sign error in Eq. (11), and the absence of an absolute calibration check for the claimed field sensitivity.","major_comments":[{"comment":"Equation (11) is inconsistent with Eq. (3): solving (Q + ∇∘B_ext|0) r0 + B_ext(0) = 0 gives r0 = −(Q + ∇∘B_ext|0)^{-1} B_ext(0), with a minus sign absent from the printed equation. This should be corrected. More importantly, the surrounding text states that inhomogeneities 'degrade' the buoy effect and that r0 'no longer simply flips sign,' which is in tension with the abstract's claim that the signal is free from weak magnetic-field inhomogeneities. For a static gradient G, a perturbative expansion of the corrected expression gives r0(Q) − r0(−Q) = −2Q^{-1} diag(1,1,−2)^{-1} B_ext(0) + O(G²/Q³), so the gradient affects the differential signal only at second order. The manuscript should state this expansion explicitly and qualify the insensitivity claim as first-order in G/Q. As written, the abstract overclaims relative to Sec. 3.3.","section":"3.3, Eq. (11)"},{"comment":"The paper demonstrates zero-crossing compensation currents but does not report the slopes of the fitted Δy vs I_y and Δz vs I_z lines, nor compare them with the Biot–Savart coefficients α_i of Eq. (5). The mG field sensitivity in Eq. (7) is derived from the nominal quadrupole gradient and pixel size, but no independent calibration point shows that a known applied field produces the predicted displacement. Without such a check, the method is convincingly demonstrated as a nulling/compensation tool, but not yet as an absolute vector magnetometer. A table of measured vs modeled slopes, or one calibration measurement against a known field, would substantiate the main sensing claim.","section":"2.2, Fig. 4"}],"minor_comments":[{"comment":"The gravity-induced centroid offset is stated as an assumption. It can be proved in one line by shifting to u = r − r0, since the potential becomes μ|Q M u| + m g u_z, independent of B_ext and of the sign of Q. Adding this argument would remove an unnecessary caveat and strengthen the differential-cancellation claim.","section":"2.4"},{"comment":"The sentence claiming that the superposed field 'retains a single magnetic zero' is only true when the matrix Q + ∇∘B_ext|0 is nonsingular. For arbitrary gradients, the zero may be non-unique or absent; the statement should be qualified to weak gradients.","section":"3.3"},{"comment":"The zero-crossing values I_y^@ = −0.27 A and I_z^@ = 0.035 A are quoted without uncertainties. Since these are the principal quantitative outputs of the demonstration, confidence intervals from the linear fits should be reported.","section":"Fig. 4"},{"comment":"The data availability statement says the data 'will be made publicly available' but does not provide a repository identifier or DOI. If possible, include the link or accession number at submission time.","section":"5"}],"recommendation":"major_revision","confidential_remarks":"The central buoy mechanism is sound, and the reported experimental trends support it. The main issues are fixable: the sign error in Eq. (11), the need for a quantitative statement about the order at which gradients enter the differential signal, and the lack of an absolute calibration check. The gravity-cancellation concern raised in the review is not a blocker. I would encourage the editor to treat this as a promising proof-of-principle that needs revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The cold-atom buoy is a simple and genuine idea: displace the center of a quadrupole trap with a homogeneous field, flip the quadrupole polarity, and the cloud-position difference gives a linear, directional readout of the external field. Eq. (3) is correct, parameter-free, and the experimental data in Figs 2–5 are internally consistent. The paper is honest about what is and isn't demonstrated—the 5 mG figure is an estimate from centroid noise, not an absolute calibration, and the 3D extension is a proposal, not a result. That is the right frame.\n\nThe one worry in the report—that gravity skews the cloud and the centroid offset might not cancel between polarities—doesn't survive contact with the actual physics. The potential under a constant gravity plus quadrupole is U(u)=μ|Q M u|+mg u_z. Reversing Q does not change |B|, so the density profile is just translated rigidly with r0; the skewness bias is polarity-independent and cancels in the difference. Similarly, a constant external gradient enters the differential signal only at second order in G/Q. So the abstract's claim about weak inhomogeneities is valid to first order. Section 3.3 makes the single-polarity response sound worse than the differential signal actually is; the paper would benefit from stating the first-order cancellation explicitly, but there is no contradiction.\n\nThe real soft spots are quantitative. There is no independent calibration of the field scale—the precision claim rests on the known quadrupole gradient and the measured centroid noise, which is fair but not a demonstration of accuracy. The raw imaging data are promised, not yet public; until they land, the matching of clusters to currents is only as convincing as the figures, which are decent. The 3D extension via an added wire is plausible but undemonstrated. None of this undermines the central mechanism.\n\nThis is a methods paper for cold-atom labs that need cheap field compensation without spectroscopy. It deserves a serious referee, not a desk reject. I'd send it out with a request for the data release and an explicit discussion of the gradient cancellation.","headline":"A clean, honest differential sensing method: Eq. (3) is right and the data fit, though the precision claim is an estimate and the 3D road map is unproven.","tokens_in":12366,"tokens_out":2397,"would_cite":false,"duration_ms":23760,"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":"Flipping the polarity of a cold-atom quadrupole trap turns the trapped cloud into a vector magnetometer sensitive to about 5 milligauss.","keywords":["cold atoms","magnetic quadrupole trap","magnetic sensing","differential metrology","absorption imaging","vector magnetometry","field compensation","milligauss resolution"],"falsifier":"Measure the two polarity-resolved cloud-center positions while varying only the atom number or temperature at a fixed external field. If the differential displacement changes with these parameters — or if the midpoint of the two clusters moves when the external bias is changed far from compensation — the assumed common-mode cancellation fails. A cleaner null test: with the compensation currents set to zero external field, the two polarity clusters should coincide; any residual separation that scales with the quadrupole gradient or gravity direction would falsify eq. (3)'s clean flip symmetry.","tokens_in":11540,"feed_emoji":"🧲","tokens_out":4822,"duration_ms":49428,"temperature":0.7,"pith_summary":"External magnetic fields displace the zero of a quadrupole magnetic trap, and reversing the trap polarity reverses that displacement. The paper turns this geometric effect into a vector magnetometer: absorption images of a trapped cold-atom cloud under both polarities give a differential displacement proportional to the external field, with no spectroscopic interrogation. In a rubidium-87 apparatus the authors demonstrate directional response to applied bias fields, locate the true magnetic center, and infer compensation currents via linear fits. With about 2-micrometer center resolution and a 2.5 G/mm gradient, the inferred field resolution is roughly 5 milligauss. The significance is practical: any cold-atom stage with a quadrupole trap and one imaging beam can measure and compensate stray fields.","feed_headline":"Trap polarity flip turns cold atoms into a 5 mG magnetometer","feed_subtitle":"No spectroscopy: comparing cloud positions under reversed quadrupole field gives a directional readout of stray fields.","key_machinery":"The machinery is the signed quadrupole field B_Q(r)=Q(x̂+ŷ−2ẑ)·r with Q proportional to coil current, together with the linear response r0 = [z∘z − 2(x∘x + y∘y)] B_ext / (2Q). A homogeneous external field shifts the trap zero along the field direction; reversing the sign of Q makes the shift opposite. The experimental implementation locates the cloud centroid by a two-dimensional Gaussian fit to the absorption image, and the differential displacement under polarity reversal is the observable. A numerical model of the coil geometry is used to calibrate the gradient and to check that deviations from the ideal quadrupole are at the microgauss level over the relevant 0.1 mm range.","core_discovery":"The central result is the identity r0 = [z∘z − 2(x∘x + y∘y)] B_ext / (2Q), where Q is the signed quadrupole strength, r0 is the displacement of the magnetic-field zero, and x, y, z are unit axes. Because the trapping potential depends only on |B|, reversing Q — by reversing the coil current — preserves the trap but flips the displacement; the midpoint of the two cloud positions then locates the true field-compensated center, while the difference is a linear vectorial measure of the external homogeneous field. The paper demonstrates the effect with 50–100 shots per condition, shows that the inferred centers cluster nearly independent of applied bias currents, and uses linear regression of the","pith_inferences":["If the 5 mG figure holds without metrological optimization, the same differential readout could be applied to atom-chip traps with much larger gradients, potentially improving resolution by an order of magnitude or more — a testable extension the paper does not pursue.","The technique could be adapted to continuously monitor field drift during long cold-atom experiments, since the atoms already in the trap can serve as the sensor between runs; this would require only reusing the standard images.","A direct experimental check of the gravity-cancellation assumption would be to vary the cloud temperature or atom number at fixed external field and see whether the polarity-resolved centroid difference (not just the mean position) shifts; if it does, an additional correction term is needed.","The buoy analogy suggests a more general metrological principle: any conservative trap whose minimum responds antisymmetrically to a controllable sign can be used as a differential field sensor, so similar schemes might work for optical or combined traps."],"forward_implications":["A single absorption-imaging axis yields both transverse components (y and z) of the external field, because each displacement component responds only to the corresponding field component.","The differential signal cancels common-mode effects such as gravity-induced centroid bias and slow drifts in coil current, leaving a direct readout of the homogeneous external field.","With current sources stable to about 1 µA, simulated common-mode current fluctuations shift the trap center by less than a micron, well below the demonstrated resolution.","The third field component can be recovered by adding a known inhomogeneous field — e.g., a current-carrying wire switched synchronously with the quadrupole polarity — coupling x-information into the imaging plane.","Because the method requires no spectroscopy or internal-state coherence, it can serve as a practical in-situ field-compensation tool in cold-atom preparation stages."],"fun_headline_variants":["Polarity flip turns cold atoms into a mG magnetometer","Differential cloud shifts cancel gravity, sense fields to mG","Reversing quadrupole current gives 3D vector magnetometry from cold atoms","No spectroscopy: cold-atom cloud positions map magnetic fields","Earth-field to atomic magnetometry via differential cold-atom shifts"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The load-bearing assumption is that the cloud centroid extracted from a symmetric 2D Gaussian fit locates the magnetic-field zero up to an offset that is identical for both quadrupole polarities; in particular, gravity skews the cloud, and the paper assumes this skew-induced displacement of the optical-density center is the same in both polarities and cancels in the differential signal.","fun_headline_variants_meta":{"raw":{"variants":["Polarity flip turns cold atoms into a mG magnetometer","Differential cloud shifts cancel gravity, sense fields to mG","Reversing quadrupole current gives 3D vector magnetometry from cold atoms","No spectroscopy: cold-atom cloud positions map magnetic fields","Earth-field to atomic magnetometry via differential cold-atom shifts"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000788,"raw_usage":{"total_tokens":3286,"prompt_tokens":694,"completion_tokens":2592,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":438,"completion_tokens_details":{"reasoning_tokens":2501}},"tokens_in":438,"tokens_out":2592,"duration_ms":19597,"temperature":1.0,"reasoning_tokens":2501,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T22:43:58.103407+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the two polarity-resolved cloud-center positions while varying only the atom number or temperature at a fixed external field. If the differential displacement changes with these parameters — or if the midpoint of the two clusters moves when the external bias is changed far from compensation — the assumed common-mode cancellation fails. A cleaner null test: with the compensation currents set to zero external field, the two polarity clusters should coincide; any residual separation that scales with the quadrupole gradient or gravity direction would falsify eq. (3)'s clean flip symmetry.","supporting_citations":[],"review_version":1}