{"id":"2e7e68d4-ac61-47a7-96c9-542e0b66a63c","arxiv_id":"2502.03334","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"A 4ħk Raman atom interferometer is demonstrated without k-reversal, using alternating microwave and Raman pulses, and its measured sensitivity n = 3.00 ± 0.05 matches prediction.","lead":"The authors demonstrated a laser and microwave pulse sequence that gives rubidium atoms a fourfold larger momentum kick in an atom interferometer without reversing the laser direction. This could simplify the design of compact, high-sensitivity accelerometers that use atom interference.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claimed n=3.00±0.05 rests on an unverified decomposition of the fringe into a main interferometer and exactly two half-area parasitic interferometers; if the parasitic weighting differs from exactly half the main trapezoid, the fitted n can be biased.","rationale":"Reading in good faith, the paper reports a genuine proof-of-principle: a 4ħk Raman LMT sequence without k-reversal, a measured fringe, and a Mach-Zehnder control that gives n_MZ=2.00±0.01, validating the acceleration readout and weighting formalism. The SM derivation of the main trapezoidal weighting is coherent and parameter-free, and the phase relation φ1=2φ2 found by the fit is a nontrivial consistency check. The load-bearing weakness is the parasitic-interferometer decomposition. It is introduced after seeing the non-sinusoidal fringes, the number and properties of the parasitic paths are asserted rather than derived, and the SM stops at the main pulse sequence without analyzing the parasitic closed paths. Because the quantitative claim n=3 is exactly the quantity that the fit is designed to extract, a misspecified model could produce a biased n. This is not an accusation of error; it is a request for a derivation or a numerical simulation that closes the gap. The short-pulse limit is a related but less critical issue, since the MZ control and the overall consistency argue that finite-pulse corrections are small at the demonstrated level. Verdict remains conditional: the proof-of-principle is credible, but the central quantitative confirmation should be backed by a derivation (or suppression) of the parasitic terms.","tokens_in":14375,"tokens_out":10344,"duration_ms":108184,"concrete_test":"Compute, with the SM path-integral method (Eq. 29), the arm-separation function x_r(t)−x_l(t) for each of the two parasitic closed paths shown in Fig. 4, including all microwave and Raman pulses with realistic non-π amplitudes. Verify that each separation is exactly one half of the main interferometer's separation at all times; if the shape differs, Eq. (5) is misspecified. Then generate synthetic fringes using the true parasitic weighting functions with a known n_main=3 and the reported noise level, fit Eq. (5), and check that the recovered n1 is unbiased. This single check settles whether the claimed 3.00±0.05 is a valid confirmation or an artifact of the assumed parasitic model.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The quantitative confirmation of the central claim—n_LMT=3.00±0.05 matching Eq. (3)—is obtained by fitting Eq. (5), which assumes that the observed non-sinusoidal fringe is the sum of the main interferometer and exactly two parasitic interferometers. The decisive assumption is not just that two such paths exist, but that each parasitic interferometer has the same trapezoidal acceleration weighting as the main one and exactly half its phase sensitivity. The SM derives the main weighting function f(t) (Eq. 25) and the general relation f(t) ∝ x_r(t) − x_l(t) (Eq. 29), but it never computes f_parasitic(t) for the two paths drawn in Fig. 4. That is a gap: a half-area path does not automatically have a half-scale trapezoid; if, for instance, one Raman pulse is skipped rather than the whole sequence being scaled, the arm-separation function can have a different shape (e.g., a longer plateau), so the parasitic phase is not strictly (1/2)Δφ_main plus a constant. A misspecified Eq. (5) can then shift the fitted n1, and the reported ±0.05 would only be the statistical error of a biased model. The short-pulse/recoil-shift issue (±15 kHz detuning against a 50 kHz Rabi coupling) affects both main and parasitic amplitudes and phases, so it is secondary to, but not independent of, this decomposition problem.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a scheme for a 4ℏk Raman large-momentum-transfer atom interferometer in which the effective Raman wavevector is not reversed. The sequence uses microwave π/2 and π pulses to swap internal states between repeated Raman π pulses, and the authors derive an acceleration sensitivity n = 4(T1/T)(2 − T1/T). They measure fringes of a proof-of-principle interferometer as a function of MEMS-tracked mirror acceleration, fit Eq. (5), and report n_LMT^(1) = 3.00 ± 0.05, together with a Mach-Zehnder control giving n_MZ = 2.00 ± 0.01. The supplement derives the trapezoidal acceleration weighting function and discusses extension to 4Nℏk.","tokens_in":14800,"tokens_out":10427,"duration_ms":106303,"significance":"The potential impact is real: eliminating k-reversal removes a frequency-switching and phase-coherence constraint in horizontal Raman LMT accelerometers, and the proposed scaling to 4Nℏk is natural. The paper has genuine strengths: the prediction n = 3 is parameter-free from the path-integral calculation; the MZ control validates the accelerometer readout at the 0.5% level; and the fitted φ1 ≈ 2φ2 and n2 ≈ 1.5 are internal consistency checks. The main weakness is that the confirmation of n = 3 rests on a two-frequency model whose parasitic weighting functions are asserted rather than derived, so the quantitative claim is not yet on the same footing as the rest of the paper.","major_comments":[{"comment":"The central result n_LMT^(1) = 3.00 ± 0.05 is obtained by fitting Eq. (5), which assumes that the observed non-sinusoidal fringe is the sum of the main interferometer and two parasitic interferometers whose acceleration phase is exactly half of the main phase with the same trapezoidal weighting. The SM derives the main weighting function f(t) in Eq. (25) and the general relation f(t) ∝ x_r(t) − x_l(t) in Eq. (29), but it does not compute x_r − x_l for the two parasitic paths drawn in Fig. 4. A path that encloses half the area does not necessarily have a half-scale copy of the main trapezoid; depending on which Raman pulse is missed, the plateau or slope times of the arm separation could differ, which would bias the fitted n1 even if the statistical error remains small. Please derive f_parasitic(t) for the parasitic paths, or alternatively add a shape parameter to the parasitic weighting in the fit and show that n1 is stable, and report a model-comparison statistic such as residuals or BIC against the half-trapezoid model.","section":"Main text, Eq. (5) and Fig. 4; SM Eqs. (25) and (29)"},{"comment":"The text states that two parasitic interferometers contribute, but Eq. (5) contains only one parasitic cosine with amplitude A2. If the two paths have the same phase sensitivity, their sum can be written as a single sinusoid only after the relative phase and amplitude of the two contributions are specified; if the two weighting functions are not identical, the single-cosine parametrization is not sufficient. The manuscript should either explain this reduction explicitly or fit the two parasitic terms separately.","section":"Eq. (5) and surrounding text, Fig. 4"},{"comment":"The Raman π pulses are 11–12 µs long and the two arms sit at opposite two-photon recoil detunings of ±2π × 15 kHz with a Rabi coupling of 2π × 50 kHz. The main derivation assumes the short-pulse limit. Please give a quantitative estimate of the residual phase error from finite pulse duration and off-resonant driving for both the main and parasitic interferometers, and state the contribution of this effect to the quoted uncertainty of ±0.05 in n1. This is needed to support the claim that the measured n1 is an unbiased test of Eq. (3).","section":"Main text, pulse parameters; SM Methods"}],"minor_comments":[{"comment":"Equation (19) states Δφ_R = a T1(T1 + T2 + T3), which has dimensions of length and differs from the main-text Eq. (3) by a factor of 4k; this is presumably a typographical omission, but the supplement should be corrected for consistency with the main text.","section":"SM Eq. (19)"},{"comment":"There are minor language errors: 'we apply we apply a Raman bias field' in the Methods and 'undertood' in the Bragg-transition section should be corrected.","section":"SM Methods and Bragg section"},{"comment":"The quantities t1, t4, t7 and the tilde times T1 and T2 are used in the supplement but are not defined in the main text or in Fig. 1; please define them explicitly in the caption or in the text.","section":"Fig. 1 and main text timing notation"},{"comment":"The MZ data were taken with the table floating while the LMT data were taken with the table not floated; the caption states this in the text, but the figure itself should make it clear that the two data sets are not taken under identical vibration conditions.","section":"Fig. 3 caption"}],"recommendation":"major_revision","confidential_remarks":"The paper is likely publishable after the authors supply a first-principles treatment or a robustness check of the parasitic model, and after fixing the units error in SM Eq. (19). The main idea is sound and the MZ control is a good validation, but the headline number 3.00 ± 0.05 is not yet supported at the level the current text claims because the extraction depends on an unverified decomposition of the fringe."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Worth a careful read. This paper demonstrates a 4ħk Raman interferometer that avoids k-reversal by alternating microwave and Raman pulses, and it does so with a clean parameter-free prediction for the acceleration sensitivity. The trapezoidal weighting function is derived properly in the supplement, and the MZ control (n=2.00±0.01) validates the experimental pipeline. The measured n=3.00±0.05 for the LMT fringe matches the predicted n=3, and the parasitic fringe with n≈1.5 and φ1≈2φ2 hangs together.\n\nThe new bit is genuine: the alternating MW/Raman sequence does remove the need for k-reversal in spin-dependent kicks, which matters for horizontal accelerometers. The path to larger momentum transfer via adiabatic passage is plausible.\n\nThe soft spot is the parasitic interferometer model. The paper explains the non-sinusoidal fringe by invoking exactly two parasitic interferometers with half the area, and the fit to Eq. (5) then yields the headline n. The supplement derives the main interferometer's weighting function and the general idea that f(t) is proportional to the arm separation, but it never computes the arm separation for the parasitic paths. So the assertion that these have exactly the same trapezoidal shape scaled by 1/2 is unproven. If the parasitic weighting is not exactly half the main trapezoid, the fitted n can shift. This is a legitimate concern, though not a red flag—the fit's internal consistency (n2=1.5, φ1=2φ2) is suggestive that the model is close.\n\nThe short-pulse approximation (recoil shifts ±15 kHz vs 50 kHz Rabi) is also not quantified, but the MZ control gives exactly n=2, so it seems benign. The fringe amplitude is small (~0.03), but the data span covers about two periods, so the fit is not nonsense.\n\nBottom line: The proof-of-principle is real, the key number matches prediction, but the confirmation is contingent on an unverified decomposition. I'd want the authors to either derive the parasitic weighting function from the path integral, or show a higher-contrast demonstration where the main fringe is unambiguous. This is a solid candidate for peer review, not a desk reject.","headline":"A credible 4ħk LMT Raman interferometer without k-reversal, with a clean prediction matched by data, but the headline n relies on an unproven parasitic-fringe model.","tokens_in":15308,"tokens_out":6474,"would_cite":true,"duration_ms":61507,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["37.25.+k"],"model":"deepseek-v4-flash","headline":"A Raman atom interferometer can achieve 4ħk momentum transfer without reversing the effective wavevector, and the measured acceleration sensitivity n = 3.00 ± 0.05 matches the predicted n = 3.","keywords":["atom interferometry","large momentum transfer","Raman transitions","spin-dependent kicks","k-reversal","87Rb","accelerometry","parasitic interferometers"],"falsifier":"Measure fringes at several $T_1/T$ values, for example 0.25, 0.5, and 0.75, and extract $n$ from each; the central claim predicts $n = 4(T_1/T)(2 - T_1/T)$, so a systematic deviation from that parabola would show the model is incomplete. A second check is to vary the Raman pulse intensity and verify that the parasitic fringe amplitude and the $\\phi_1 = 2\\phi_2$ relation move together as the two-parasitic-interferometer model requires.","tokens_in":14211,"feed_emoji":"⚛️","tokens_out":11162,"duration_ms":89754,"temperature":0.7,"pith_summary":"Large-momentum-transfer (LMT) atom interferometers normally need to reverse the direction of the effective Raman wavevector between light pulses. This paper shows how to obtain a $4\\hbar k$ momentum transfer without that reversal, by inserting microwave pulses that swap the two internal spin states while leaving momentum unchanged. The proof-of-principle measurement gives an acceleration sensitivity of $n = 3.00 \\pm 0.05$, matching the predicted value $n = 4(T_1/T)(2 - T_1/T) = 3$ for the timing used. This matters because the sequence removes a technical constraint on horizontal and compact atom accelerometers without sacrificing the space-time area gain of large momentum transfer.","feed_headline":"No-reversal atom interferometer hits 4ħk momentum transfer","feed_subtitle":"Measured acceleration sensitivity of 3.00 matches the predicted value, easing a key constraint on compact atom accelerometers.","key_machinery":"The load-bearing element is the alternating microwave/Raman pulse sequence, where microwave pulses switch the internal spin state without changing momentum, so the same Raman wavevector can open and close the interferometer arms. The quantitative carrier is the dimensionless sensitivity factor $n = 4(T_1/T)(2 - T_1/T)$ together with the trapezoidal acceleration weighting function, which is proportional to the time-dependent separation of the two arms and replaces the triangular weighting of a standard Mach-Zehnder interferometer.","core_discovery":"Starting from a microwave $\\pi/2$ pulse that puts each atom in an equal superposition of the two hyperfine states, the sequence uses two Raman $\\pi$ pulses to separate the arms by $4\\hbar k$, a central microwave $\\pi$ pulse to swap the spin states of the two arms, and two more Raman $\\pi$ pulses to close the interferometer without reversing the wavevector. With symmetric timing, the Raman phase is $\\Delta\\phi_R = a\\, n\\, k\\, T^2$ with $n = 4(T_1/T)(2 - T_1/T)$, and the mirror-acceleration weighting is trapezoidal rather than triangular. The measured fringe gives $n = 3.00 \\pm 0.05$ for $T_1/T = 1/2$, confirming the factor-of-1.5 spacing improvement over a three-pulse Mach-Zehnder interferometer ($n = 2.00 \\pm 0.01$). The observed non-sinusoidal fringe is accounted for by two parasitic interferometers that enclose half the area, with fitted sensitivity $n = 1.50 \\pm 0.04$ and phase relation $\\phi_1 = 2\\phi_2$.","pith_inferences":["Beyond the paper: if the two-parasitic-interferometer decomposition is correct, the ratio of parasitic to main fringe amplitude should vary systematically with Raman $\\pi$-pulse fidelity, so deliberately tuning the pulse intensity would provide an independent test.","Beyond the paper: because no intra-sequence $k$-reversal is needed, the sign of $k$ can be flipped between successive measurements, and differencing the two fringe phases should cancel any acceleration-independent phase offset.","Beyond the paper: the $n$ versus $T_1/T$ curve is a direct prediction of the model, so mapping fringes at several $T_1/T$ values would test the theory beyond the single operating point reported here."],"forward_implications":["A $4\\hbar k$ Raman interferometer can be operated without $k$-reversal, with measured acceleration sensitivity $n = 3.00 \\pm 0.05$ at $T_1/T = 1/2$.","For the same total interrogation time, the fringe spacing is 1.5 times finer than a three-pulse Mach-Zehnder interferometer.","Mirror-vibration noise must be averaged with a trapezoidal weight over this sequence, and the same geometric argument extends the weighting to arbitrary pulse sequences.","The parasitic interferometers are identifiable and benign: they enclose half the area and obey $\\phi_1 = 2\\phi_2$, so the main fringe can be recovered by fitting.","The alternating sequence scales to $4N\\hbar k$ using $(4N-1)$ microwave and $4N$ Raman pulses, and with roughly 96% pulse efficiency the paper expects $16\\hbar k$ at contrast similar to earlier spin-dependent-kick work."],"supporting_citations":[{"why":"Supplies the spin-dependent-kick LMT technique and the adiabatic-passage pulse-efficiency estimate that the proposed $4N\\hbar k$ extension relies on.","marker":"[34]"},{"why":"Earlier LMT Raman interferometer that required $k$-reversal; the present scheme is the no-reversal alternative.","marker":"[37]"},{"why":"Provides the apparatus, Raman laser system, and measurement environment used for the demonstration.","marker":"[40]"},{"why":"Gives the short-pulse-limit phase calculation used to derive the acceleration phase and the sensitivity factor $n$.","marker":"[26]"},{"why":"Supplies the propagation/interaction phase decomposition used in the supplemental derivation of the total phase.","marker":"[44]"},{"why":"Provides the path-integral argument that identifies the acceleration weighting function with the time-dependent arm separation.","marker":"[46]"},{"why":"Provides the vibration-weighted phase analysis for Mach-Zehnder interferometers that the paper adapts to a trapezoidal weighting.","marker":"[47]"},{"why":"Documents the MEMS accelerometer used to measure the mirror acceleration that defines the fringe axis.","marker":"[48]"}],"fun_headline_variants":["4ħk atom interferometer demonstrated without k-reversal","Microwave pulses enable 4ħk LMT without k-reversal","Proof-of-principle: atom interferometer skips k-reversal","Raman interferometer achieves 4ħk transfer via microwave swaps","No-reversal design yields 4ħk momentum transfer in interferometer"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The central claim rests on the fit model that the observed non-sinusoidal fringe is the sum of the main interferometer and exactly two parasitic interferometers enclosing half its area with phase relation $\\phi_1 = 2\\phi_2$; if that decomposition is wrong, the quoted $n = 3.00$ does not follow.","fun_headline_variants_meta":{"raw":{"variants":["4ħk atom interferometer demonstrated without k-reversal","Microwave pulses enable 4ħk LMT without k-reversal","Proof-of-principle: atom interferometer skips k-reversal","Raman interferometer achieves 4ħk transfer via microwave swaps","No-reversal design yields 4ħk momentum transfer in interferometer"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000336,"raw_usage":{"total_tokens":1855,"prompt_tokens":932,"completion_tokens":923,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":548,"completion_tokens_details":{"reasoning_tokens":831}},"tokens_in":548,"tokens_out":923,"duration_ms":8157,"temperature":1.0,"reasoning_tokens":831,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-09T05:04:59.811934+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure fringes at several $T_1/T$ values, for example 0.25, 0.5, and 0.75, and extract $n$ from each; the central claim predicts $n = 4(T_1/T)(2 - T_1/T)$, so a systematic deviation from that parabola would show the model is incomplete. A second check is to vary the Raman pulse intensity and verify that the parasitic fringe amplitude and the $\\phi_1 = 2\\phi_2$ relation move together as the two-parasitic-interferometer model requires.","supporting_citations":[{"cited_title":"Jaffe, V","cited_arxiv_id":null,"evidence_quote":"Supplies the spin-dependent-kick LMT technique and the adiabatic-passage pulse-efficiency estimate that the proposed $4N\\hbar k$ extension relies on."},{"cited_title":"McGuirk, M","cited_arxiv_id":null,"evidence_quote":"Earlier LMT Raman interferometer that required $k$-reversal; the present scheme is the no-reversal alternative."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the apparatus, Raman laser system, and measurement environment used for the demonstration."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the propagation/interaction phase decomposition used in the supplemental derivation of the total phase."},{"cited_title":"Storey and C","cited_arxiv_id":null,"evidence_quote":"Provides the path-integral argument that identifies the acceleration weighting function with the time-dependent arm separation."},{"cited_title":"Barrett, P.-A","cited_arxiv_id":null,"evidence_quote":"Provides the vibration-weighted phase analysis for Mach-Zehnder interferometers that the paper adapts to a trapezoidal weighting."},{"cited_title":"Beitia, A","cited_arxiv_id":null,"evidence_quote":"Documents the MEMS accelerometer used to measure the mirror acceleration that defines the fringe axis."}],"review_version":1}