{"id":"2426c2e8-6b9e-4e1b-bf02-28d6ec1d3f34","arxiv_id":"2506.04315","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Entangling a transmon qubit with a positron-like antiqubit whose Hamiltonian is inverted achieves the maximum possible Fisher information per space-time volume for phase estimation with an unknown field direction.","lead":"This paper makes a special type of superconducting qubit, called an antiqubit, that behaves like a positron: its response to a magnetic field is reversed. Entangling an ordinary qubit with an antiqubit lets the pair measure the strength of a field of unknown direction with the best possible sensitivity per resource, a result they prove and demonstrate in the lab.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Experimental arbitrary-direction inversion hinges on a magic AC-Stark frequency that is calibrated from data, not predicted; the 9.52 MHz detuning from the qubit transition causes residual off-resonant rotations that could reduce the z-axis Fisher information below the claimed average.","rationale":"The reader's weakest-assumption diagnosis points to the same experimental vulnerability, and I agree that it is the most load-bearing concern for the paper's full claim. The theoretical proof of the FI=4 bound and the uniqueness of the singlet-plus-inversion strategy appears sound: the bound I ≤ 2[1+C] in Suppl. Note I C is proven before the uniqueness argument in Note I D, so the use of Eq. (26) there is justified by the preceding maximal-entanglement result. The experimental demonstration, however, depends on a magic AC-Stark frequency whose location is inferred from data rather than predicted, and whose proximity to the qubit transition introduces known off-resonant rotations. Because the paper's abstract and title assert that superconducting antiqubits 'achieve optimal phase estimation' and that 'we observe it experimentally,' the z-inversion calibration is not a peripheral detail. The paper is transparent about the discrepancy and the contamination, which is good scientific practice, but the transparency also reveals that the arbitrary-direction unitary inversion has not been validated from first principles. A clean, independent prediction of the magic frequency would settle whether the z-inversion is a genuine physical effect or a calibrated artifact. Since the reader already marked the paper CONDITIONAL and this concern reinforces that condition rather than moving the verdict, the recommended disposition remains conditional acceptance.","tokens_in":28759,"tokens_out":20901,"duration_ms":199453,"concrete_test":"Measure the drive amplitude ratio R = Ω_antiqubit/Ω_qubit independently using Rabi oscillations at a common detuning, and use Eq. (89) with the independently characterized ω and α from Table I to predict the magic frequency with no free parameters. If the prediction does not match 4.176998 GHz within the Ramsey linewidth, the magic-frequency condition is empirical rather than understood from first principles. Separately, characterize the residual off-resonant Rabi rate at the magic frequency by applying the Stark tone to a qubit prepared in |+x> and measuring the induced rotation; if this residual rotation per target z-rotation exceeds ~0.1 rad across the α range, the z-axis FI inferred from the P(|Ψ−>) ≈ cos²(α) curve is unreliable and the average FI should be recomputed excluding or correcting the z-axis data.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The experimental leg of the central claim—arbitrary-direction unitary inversion and the resulting FI of 3.02 per v_st = 2—rests on the z-component inversion being accurate. In Sec. IV and Supplementary Note III, the magic frequency is inferred from Ramsey measurements (4.176998 GHz) rather than predicted from first principles; the predicted value is 4.19742 GHz, and the discrepancy is attributed to a 1.78x field-amplitude difference that is not independently measured. Moreover, the inferred magic frequency is only 9.52 MHz detuned from the qubit transition, so the Stark tone also drives off-resonant x/y rotations. The paper acknowledges that this contaminates the z-axis data (Fig. 3(d) and Suppl. Note III). If the magic-frequency condition is not robust or if residual off-resonant rotations are significant, the z-axis FI could be substantially lower than the ideal cos²(α) value, and the reported average FI of 3.02 ± 0.07 would not support the claim of experimentally realizing optimal phase estimation for an arbitrary field direction. This concern does not touch the theoretical optimality proof, but it directly affects the experimental demonstration, which is a central part of the paper's contribution.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes and demonstrates \"antiqubits\": transmons driven so that their effective gyromagnetic ratio is opposite to that of an ordinary qubit, thereby realizing platform-specific unitary inversion. The central theoretical claim is that an entangled qubit–antiqubit singlet, with the field acting as U_α on the qubit and U†_α on the antiqubit, achieves the maximal Fisher information per two units of space–time volume, I_α = 4, and that this strategy is unique for unknown field direction. The experimental section reports P(|Ψ−⟩) ≈ cos(2α) curves, an averaged Fisher information of 3.03 ± 0.07 per v_st = 2, and comparisons with separable qubit–antiqubit and agnostic-sensing strategies. The supplement contains the QFI derivation, the uniqueness proof, and the experimental calibration details, including the AC-Stark \"magic frequency\" used for z-axis inversion.","tokens_in":29038,"tokens_out":7250,"duration_ms":73077,"significance":"If the proof is completed and the experimental z-axis data are robust, this is a valuable contribution: it offers a physical, low-overhead route to unitary inversion and demonstrates a clear metrological advantage in a platform-specific setting. The paper's strengths include an explicit analytic derivation of the QFI bound, a self-contained supplement with derivations of Eq. (13) and the concurrence bound, careful resource accounting via the space–time volume, and direct experimental comparison with two competitor strategies. The uniqueness claim, once repaired, would be a strong and interesting result that goes beyond merely showing an advantage.","major_comments":[{"comment":"The uniqueness proof applies Eq. (26) to a general pure two-TLS state, but Eq. (26) was derived in Supplementary Note I B under the assumption that the state is maximally entangled (vanishing Bloch vectors). For a general pure state, the correct QFI expression is Eq. (13), which contains the additional nonpositive term −(r_A_n + s r_B_n)². As written, the step \"the optimal ˆn-independent strategy implies, by Eq. (26), that I(s)_α = 2(1 + s n^T T n) = 4\" is therefore not justified for general |ψ⟩. This gap is load-bearing because it underlies the claim that the singlet plus effective unitary inversion is the unique optimal strategy. The gap is likely repairable, e.g., by first using the bound I_α ≤ 2[1 + C(|ψ⟩)] ≤ 4 to conclude that any strategy reaching I_α = 4 for all ˆn must have C = 1, and only then applying Eq. (26), but the current text needs this argument spelled out.","section":"Supplementary Note I D"},{"comment":"The main text states that P(|Ψ−⟩) ≈ cos(2α) (Fig. 3(b) and surrounding text), while Supplementary Note I E, Eq. (70), derives P(|Ψ−⟩) = cos²(α) for the same protocol. These two expressions are not equivalent: cos(2α) takes negative values on (π/4, 3π/4), which is impossible for a probability, and the FI formula in Eq. (71)–(72) is computed from cos²(α). The two statements should be reconciled; presumably the intended probability is cos²(α) = [1 + cos(2α)]/2, with the \"2\" in the text referring to the doubled effective angle under U²_α rather than to a cos(2α) functional form. As written, the experimental claim that the curve supports the effective-U² interpretation is internally inconsistent with the theoretical derivation.","section":"Sec. V and Supplementary Note I E"},{"comment":"The experimental claim of arbitrary-direction unitary inversion rests on the z-component inversion produced by the AC-Stark magic frequency. The supplement reports that the inferred magic frequency, 4.176998 GHz, differs from the predicted 4.19742 GHz by a factor attributed to a 1.78× field-amplitude difference that is not independently measured, and that the inferred frequency is only 9.52 MHz detuned from the qubit transition, causing off-resonant x/y rotations that visibly contaminate the z-axis data (\"gray disks ... wobble more\"). The paper does not report the per-axis Fisher information values from Fig. 3(e), so it is not possible to assess whether the z-axis FI is substantially below the average of 3.03 ± 0.07 or whether the average is carried by the x- and y-axes. Since the central experimental message is that the optimal FI is achieved for an arbitrary (unknown) field direction, the authors should report the three per-axis FIs with uncertainties and quantify the sensitivity of the z-axis FI to the magic-frequency calibration and to the residual off-resonant rotations.","section":"Sec. IV, Fig. 2(e), and Supplementary Note III"}],"minor_comments":[{"comment":"The abstract reports an experimental FI of 3.02 per two units of space–time volume, while Sec. V reports 3.03 ± 0.07; the two numbers should be made consistent.","section":"Abstract vs. Sec. V"},{"comment":"The phrase \"per two units of phase-space volume\" appears in the text of Sec. V; elsewhere the paper consistently uses \"space–time volume.\" The typo should be corrected.","section":"Sec. V"},{"comment":"The main text quotes the magic frequency as 4.177 GHz, while the supplement quotes 4.176998 GHz; the text should state explicitly that the former is a rounded value, and ideally give both the predicted and inferred magic frequencies in the main text with a brief explanation of the 1.78× amplitude discrepancy.","section":"Fig. 2(e) and Supplementary Note III"},{"comment":"The extension to arbitrary space–time volume is cited as reference [45] (\"in prep.\"); the statement that this achieves \"the optimal value\" should be clearly marked as an outlook result based on unpublished work rather than as a theorem proved in the present manuscript.","section":"Sec. VI"},{"comment":"The proof of uniqueness should explicitly state that \"optimal\" means achieving I_α = 4 for every field direction ˆn, not merely on average over ˆn, since this is what the subsequent algebraic condition T = s1 uses.","section":"Supplementary Note I D"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is well written and the core idea is attractive, but the uniqueness proof has a repairable gap and the experimental z-axis claim needs per-axis numbers. The cos(2α) versus cos²(α) inconsistency, together with the 3.02/3.03 discrepancy, suggests the manuscript needs a careful consistency pass. I would be willing to review a revised version."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a solid paper that proves a clean optimality claim and backs it with a credible experiment. The new idea is the antiqubit — a transmon whose effective gyromagnetic ratio is negated by Z gates and an AC Stark drive — which lets them implement U† on one member of an entangled pair. Combined with a singlet, that yields Fisher information 4 per two units of space-time volume, and they prove this is the unique optimum for unknown-axis phase estimation at vst=2. That uniqueness proof is the main theoretical contribution, and it is mostly rigorous.\n\nThe experiment is the real thing: they prepare the singlet with 97% fidelity, implement inversion for x, y, and z axes, and get P(|Ψ−⟩) ≈ cos²α. The reported FI of 3.02 ± 0.07 per vst=2 is consistent with state-preparation and measurement limitations. The comparison with the separable qubit–antiqubit protocol is fair and shows that entanglement plus inversion is doing the work.\n\nSoft spots, in proportion. First, the uniqueness proof in Supplementary Note I.D has a repairable gap: Eq. (26) — the formula for a maximally entangled state's QFI — is applied to general pure states before maximal entanglement is established. The later concurrence analysis fixes the logic, but as written the reader has to fill in a step. Second, the magic AC Stark frequency at which δq = −δq̄ is inferred from data (4.176998 GHz) rather than predicted; the predicted value is 4.19742 GHz, and the discrepancy is blamed on a 1.78× field-amplitude difference that isn't independently measured. The paper acknowledges that the near-resonant tone causes off-resonant x/y rotations that visibly contaminate the z-axis data. That is a real limitation of the experimental demonstration, but it does not touch the theoretical optimality proof, and the reported average FI still beats the competitors by a large margin. Third, data and code are not public — for an experimental claim this strong, that is a legitimate request for revision. The multi-application extension is cited to an unpublished manuscript; that is minor since the vst=2 claim stands on its own.\n\nThe citation pattern is fine: they cite their own prior agnostic phase estimation, which is the direct antecedent, plus the relevant unitary-inversion and metrology literature. Nothing circular.\n\nWho this is for: quantum metrology people and experimentalists working on superconducting circuits. It deserves a serious referee — the proof gap is fixable, and the experiment is worth scrutiny. I would send it out, with a request for the missing derivation step and ideally the data.","headline":"A clean proof of optimal FI=4 for agnostic phase estimation via 'antiqubit' unitary inversion, with a credible but imperfect experiment; the main gap is a fixable step in the uniqueness proof.","tokens_in":29576,"tokens_out":1841,"would_cite":true,"duration_ms":19708,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81P50","81P68","81P40"],"pacs":[],"model":"deepseek-v4-flash","headline":"An entangled qubit and antiqubit sensor reaches the maximum possible Fisher information, 4 per two units of space-time volume, for measuring the strength of a field pointing in an unknown direction; the experiment records about 3.0.","keywords":["antiqubit","unitary inversion","phase estimation","Fisher information","space-time volume","superconducting transmon","quantum metrology","entangled sensing"],"falsifier":"A direct check is to measure $\\delta_q$ and $\\delta_{\\bar{q}}$ simultaneously with Ramsey interferometry while sweeping drive amplitude at the claimed magic frequency 4.176998 GHz; if their ratio departs from $-1$ by more than the linewidth, arbitrary-direction unitary inversion fails. Equivalently, fit the singlet-survival probability $P(|\\Psi^-\\rangle)$ versus $\\alpha$ for a field axis that maximizes the residual off-resonant rotation: the theory demands Fisher information 4 at the optimal $\\alpha$ for every axis, so an inferred value clearly below 4 after readout correction would refute the claim.","tokens_in":28596,"feed_emoji":"⚛️","tokens_out":15120,"duration_ms":136063,"temperature":0.7,"pith_summary":"The paper claims that a transmon can be dressed to behave like a positron, an \"antiqubit\" with an effective gyromagnetic ratio opposite to a normal qubit's, so that a magnetic field applies a unitary $U_\\alpha$ to the qubit and its inverse $U_\\alpha^\\dagger$ to the antiqubit. This platform-specific unitary inversion lets a singlet-entangled qubit-antiqubit pair measure the strength of a field pointing in an unknown direction with Fisher information 4 per two units of space-time volume, the greatest amount possible for that resource budget. The paper proves that entanglement together with effective unitary inversion is the unique strategy reaching that bound, and reports an experimental Fisher information of $3.03 \\pm 0.07$ per two units of space-time volume. The practical payoff is that unknown-direction phase estimation no longer needs to know or learn the field axis, and the inversion costs no extra applications of the unitary.","feed_headline":"Entangled qubit and antiqubit hit sensing limit","feed_subtitle":"A transmon made positron-like lets two entangled qubits reach the theoretical sensing limit.","key_machinery":"The load-bearing mechanism is platform-specific unitary inversion, realized by negating the effective gyromagnetic ratio of one transmon. Z pulses reverse the x- and y-components of the perceived field, while an off-resonant drive at the magic frequency 4.177 GHz produces an AC Stark shift (a drive-induced energy shift) with equal and opposite values on the two transmons, $\\delta_q = -\\delta_{\\bar{q}}$, reversing the z-component. When combined with the singlet state, the joint evolution $(U_\\alpha \\otimes U_\\alpha^\\dagger)|\\Psi^-\\rangle$ is equivalent to evolving one probe by $U_\\alpha^2$, so the generator's effective spectral gap doubles and the Fisher information quadruples. The resource metric is the space-time volume, (number of transmons) times (number of sequential applications of the unitary), fixed to $v_{\\rm st}=2$ for the main proof and generalized to Fisher information $4n^2$ for $n$ sequential applications.","core_discovery":"The central discovery is that antimatter's time-reversal property can be simulated on a superconducting transmon and turned into a metrological resource. Two control techniques give the \"antiqubit\" an effective magnetic response opposite to the qubit's: Z gates reverse the x- and y-components of the field, and an off-resonant drive at a magic frequency gives the two transmons equal-and-opposite energy shifts, reversing the z-component. Under a field in any direction, the qubit evolves by $U_\\alpha$ while the antiqubit evolves by $U_\\alpha^\\dagger$; because the singlet $|\\Psi^-\\rangle$ is invariant under $U_\\alpha \\otimes U_\\alpha$, the pair's evolution is equivalent to one probe evolving under $U_\\alpha^2$, doubling the phase accumulated. This yields Fisher information 4 per two units of space-time volume, matching the quantum Fisher-information bound, and the paper proves that any strategy attaining that bound must use a maximally entangled state and effective unitary inversion. The experimental singlet-survival curves give FI $= 3.03 \\pm 0.07$ per two units of space-time volume, versus 1 for two entangled ordinary qubits and $4/3$ for a separable qubit-antiqubit pair.","pith_inferences":["The observed 1.78-fold field-amplitude asymmetry between qubit and antiqubit suggests the magic-frequency condition depends on drive and wiring asymmetries; engineering those asymmetries predictably should let the experimental FI climb from about 3.0 toward 4, and would remove the need to infer the magic frequency from data.","Because the inversion flips the sign of the effective generator on all three axes at once, the same pair could serve a multiparameter problem, estimating field strength and direction simultaneously; the paper does not develop this, but its quantum-Fisher-information-matrix analysis points to it.","The effective-$U^2$ equivalence implies the sensitivity boost comes from sign-flipping the coupling rather than from adding photons, qubits, or sequential applications; any platform with a controllable sign flip of its Hamiltonian could in principle reproduce the improvement."],"forward_implications":["An unknown-direction field can be measured at the quantum Fisher-information limit without adaptive measurements, because the singlet plus inverted gyromagnetic ratio removes the need to know the rotation axis.","Repeated applications extend the advantage: $n$ sequential field applications on one synthetic-positronium pair give Fisher information $4n^2$, the quantum Fisher information, and an optimal FI per unit space-time volume of $2n$.","Platform-specific unitary inversion supplies $U^\\dagger$ at the cost of a drive configuration rather than the $O(d^2)$ applications needed by tomography-based reversal, so algorithms that call both $U$ and $U^\\dagger$ could inherit the saving.","True positronium would implement the same protocol with no control overhead, because the positron's gyromagnetic ratio exactly mirrors the electron's; readout is available through annihilation-lifetime spectroscopy, and longer-lived triplet states can support related entanglement advantages."],"supporting_citations":[{"why":"Gives the positron-as-electron-moving-backward-in-time picture that motivates simulating antimatter for unitary inversion.","marker":"[1-4]"},{"why":"Defines the space-time volume resource metric used to compare Fisher information per sensor per sequential unitary application.","marker":"[30]"},{"why":"Supplies the agnostic phase-estimation protocol and singlet-probe measurement scheme that positronium metrology extends, and the previous two-qubit baseline.","marker":"[35]"},{"why":"Contains the proof that Fisher information 4 equals the quantum Fisher-information bound and that entanglement plus effective unitary inversion is the unique optimal strategy.","marker":"[38]"},{"why":"In-preparation extension showing that $n$ sequential field applications on one qubit-antiqubit pair reach Fisher information $4n^2$.","marker":"[45]"},{"why":"Establishes the annihilation-lifetime spectroscopy readout that would let true positronium implement the protocol's measurement.","marker":"[46]"},{"why":"Supplies the concurrence measure entering the bound $I_\\alpha \\le 2[1+C(|\\psi\\rangle)]$ that ties entanglement to achievable quantum Fisher information.","marker":"[67]"},{"why":"Provides the energy-shift formula used to predict and locate the magic frequency where the qubit and antiqubit shifts are equal and opposite.","marker":"[73]"}],"fun_headline_variants":["Antiqubit twin hits quantum sensing limit","Time-reversed transmon boosts phase precision","Qubit and antiqubit pair reaches FI bound","Simulated positron doubles sensing sensitivity"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that a single microwave frequency can make the qubit and antiqubit experience exactly equal and opposite energy shifts while the residual off-resonant rotation it also causes stays small; that magic frequency is inferred from calibration data rather than predicted from the device model, and it sits only 9.52 MHz from the qubit transition, where the unwanted rotation visibly contaminates the z-axis data.","fun_headline_variants_meta":{"raw":{"variants":["Antiqubit twin hits quantum sensing limit","Time-reversed transmon boosts phase precision","Qubit and antiqubit pair reaches FI bound","Simulated positron doubles sensing sensitivity"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000142,"raw_usage":{"total_tokens":1182,"prompt_tokens":971,"completion_tokens":211,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":587,"completion_tokens_details":{"reasoning_tokens":155}},"tokens_in":587,"tokens_out":211,"duration_ms":2856,"temperature":1.0,"reasoning_tokens":155,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T10:45:42.037113+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct check is to measure $\\delta_q$ and $\\delta_{\\bar{q}}$ simultaneously with Ramsey interferometry while sweeping drive amplitude at the claimed magic frequency 4.176998 GHz; if their ratio departs from $-1$ by more than the linewidth, arbitrary-direction unitary inversion fails. Equivalently, fit the singlet-survival probability $P(|\\Psi^-\\rangle)$ versus $\\alpha$ for a field axis that maximizes the residual off-resonant rotation: the theory demands Fisher information 4 at the optimal $\\alpha$ for every axis, so an inferred value clearly below 4 after readout correction would refute the claim.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the space-time volume resource metric used to compare Fisher information per sensor per sequential unitary application."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the agnostic phase-estimation protocol and singlet-probe measurement scheme that positronium metrology extends, and the previous two-qubit baseline."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Contains the proof that Fisher information 4 equals the quantum Fisher-information bound and that entanglement plus effective unitary inversion is the unique optimal strategy."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"In-preparation extension showing that $n$ sequential field applications on one qubit-antiqubit pair reach Fisher information $4n^2$."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the annihilation-lifetime spectroscopy readout that would let true positronium implement the protocol's measurement."},{"cited_title":"Wootters","cited_arxiv_id":null,"evidence_quote":"Supplies the concurrence measure entering the bound $I_\\alpha \\le 2[1+C(|\\psi\\rangle)]$ that ties entanglement to achievable quantum Fisher information."},{"cited_title":"Carroll, S","cited_arxiv_id":null,"evidence_quote":"Provides the energy-shift formula used to predict and locate the magic frequency where the qubit and antiqubit shifts are equal and opposite."}],"review_version":1}