Superconducting-antiqubit simulation enables optimal phase estimation via unitary inversion

arXiv:2506.04315 · quant-ph, cond-mat.mes-hall, hep-ex, physics.atom-ph · Submitted 2025-06-04 · Read on arXiv

Listen

Transcript

Introduction to the show: ident: Quantum Radio. Generated commentary on the latest quantum physics and condensed matter papers.

Kai: Today's paper: "Superconducting-antiqubit simulation enables optimal phase estimation via unitary inversion".

Mira: The gist: Positronium metrology achieves an FI of 4 while consuming vst = 2 units of space–time volume, which equals the greatest possible QFI achievable with any strategy,

Kai: First, who's behind it and why it matters.

Title and authors: Kai: So, to recap, this paper shows that by combining entanglement with a specific way of inverting unitaries—what they call platform-specific unitary inversion—you can get an information gain of four while only using two units of space-time volume #pg1.

Mira: Right. It’s about getting the absolute theoretical maximum performance for phase estimation, and they found it happens when you have this specific singlet state between the qubits #pg1.

Kai: That's the main point, but it’s not just about hitting a number; it’s that this works even if you don't know which way the external field is pointing #pg1.

Mira: Exactly. The theory says this setup lets you measure phase estimation without needing to know the field direction at all, which is pretty big for real-world sensing applications #pg7.

Kai: It means we can design sensors that are robust against orientation errors, instead of needing those complex directional measurements beforehand #pg16.

Lev: From my side, I’m thinking about how hard it is to actually keep that singlet state stable and then perform that inversion reliably on noisy hardware #pg18.

Mira: That's the practical hurdle we need to think about—the fidelity required for this whole sequence of operations needs to be really high for this advantage to matter #pg18.

Kai: Yeah, and they did show a comparison where their method beats entanglement-free protocols by a lot, achieving an FI of four versus something like one point three #pg7.

Mira: That gap between the two results really shows that the combination of entanglement and inverting those unitaries is what actually unlocks that superior sensing capability #pg7.

Kai: So if we can build these antiqubit systems with enough control, we could get sensing performance far beyond what we see with just standard entangled qubits for the same amount of hardware #pg18.

Mira: It opens up a whole new way to think about using simulated particles in quantum sensors, even before we have actual antimatter in the lab #pg18.

Kai: And they’re hinting that this unitary inversion trick isn't just for sensing; it could be useful for things like breaking cryptographic schemes down or accessing black-box systems too #pg16.

Mira: It sounds like the underlying physics of time reversal in these systems might have some real computational implications beyond just improving how well we measure a field #pg16.

Kai: So the paper lays out the perfect theoretical condition—singlet state, Uα on one, U†α on the other—and then shows how their experimental setup realizes that for this specific QFI of four #pg18.

Mira: The big takeaway is that if you can control these simulated particles enough to perform those precise unitary inversions, you get a unique optimal sensing strategy that hits the absolute resource limit #pg18.

The paper's summary: Kai: So, we're looking at what they suggest as next steps for this work on simulated antimatter and phase estimation.

Mira: They are pointing toward extending this protocol to get even better information gain per unit of space-time volume than that four we saw before #pg1.

Kai: That means trying to push past the limit of four, which is what they call the maximum QFI achievable with two units of space-time volume #pg1.

Lev: From an error correction standpoint, trying to scale up the number of qubits while maintaining that high fidelity for those unitary inversions is going to be a massive engineering challenge #pg18.

Mira: It requires finding ways to make those simulations more precise so we can probe even deeper into that theoretical bound without losing too much signal #pg18.

Kai: They're also suggesting this technique could be used for other kinds of quantum computation, like simulating complex quantum channels or even doing black-box access algorithms #pg16.

Mira: That’s a big idea because it implies that the way we simulate these exotic particles might be useful for tackling more general problems in AI and computing, not just sensing #pg16.

Kai: So if they can get better than four, what does that actually mean for how we use quantum hardware in practical applications right now?

Lev: It means we’re still dealing with the reality of noise and decoherence; getting closer to those theoretical limits requires much cleaner systems than what we have today #pg18.

Mira: The authors are also hinting that this approach could eventually lead us to explore using actual antimatter in these sensing protocols, though they say that's a way off for now #pg18.

Kai: So the immediate focus seems to be on refining the simulation itself and seeing if we can get closer to that four-unit limit with the current hardware constraints #pg1.

The paper's improvements: Kai: So we’re wrapping up on "Superconducting-antiqubit simulation enables optimal phase estimation via unitary inversion" by Song and colleagues.

Mira: Essentially, they showed that combining entanglement with those platform-specific unitary inversions lets you get the absolute best possible information gain, which is four units of Fisher information using only two units of space-time volume #pg1.

Kai: It’s a neat result because it achieves that maximum QFI even when you don't know the direction of the external field #pg1.

Lev: From my side, I still have to stress that getting those exact singlet states and then implementing the necessary drive control for the inversion reliably is going to be a very tough hurdle for any real hardware we try to build.

Mira: That’s fair, Lev. The paper lays out the perfect conditions—the specific input state and the sequence of operations—that lead directly to that four information gain number #pg18.

Kai: And it shows that this isn't just a theoretical trick; they actually built the setup with three transmons, a qubit, an antiqubit, and a coupler #pg5.

Lev: I guess seeing the experimental realization helps ground the theory, but we need to see how robust that system is when you start adding more components or trying to run error correction codes on it #pg18.

Mira: The paper does mention that this method isn't just for metrology; they suggest it could be useful for things like black-box access algorithms and breaking cryptographic schemes #pg16.

Kai: That’s a big implication because it suggests that the physics of time reversal in these systems might have applications way beyond just measuring fields #pg16.

Lev: I agree, if you can do effective time reversal on the evolution, you could potentially design new types of resilient quantum codes or algorithms for error correction #pg16.

Mira: So, while they haven't touched on scaling up the FI per unit of space-time volume past four yet, it sets a very high benchmark for what’s possible with these simulated particles #pg18.

Kai: Yeah, they want to explore that next step of getting even more information from the same amount of hardware we're using #pg1.

Lev: For now, I think the immediate focus needs to be on proving that this protocol works with enough fidelity on noisy chips before we worry about pushing past the four-unit limit #pg18.

Mira: Well, "Superconducting-antiqubit simulation enables optimal phase estimation via unitary inversion" definitely shows how entanglement and inversion combine for unique sensing power #pg18.

Conclusion: Kai: So we're wrapping up on "Superconducting-antiqubit simulation enables optimal phase estimation via unitary inversion" by Song and colleagues today.

Mira: Basically, they showed that combining entanglement with platform-specific unitary inversions lets you get an information gain of four while only using two units of space-time volume.

Kai: It’s a neat result because it achieves that maximum QFI even when you don't know the direction of the external field.

Mira: That’s fair, Lev. The paper lays out the perfect conditions—the specific input state and the sequence of operations—that lead directly to that four information gain number.

Kai: And it shows that this isn't just a theoretical trick; they actually built the setup with three transmons, a qubit, an antiqubit, and a coupler.

Mira: The paper does mention that this method isn't just for metrology; they suggest it could be useful for things like black-box access algorithms and breaking cryptographic schemes.

Kai: That’s a big implication because it suggests that the physics of time reversal in these systems might have applications way beyond just measuring fields.

Lev: I agree, if you can do effective time reversal on the evolution, you could potentially design new types of resilient quantum codes or algorithms for error correction.

Mira: So, while they haven't touched on scaling up the FI per unit of space-time volume past four yet, it sets a very high benchmark for what’s possible with these simulated particles.

Kai: Yeah, they want to explore that next step of getting even more information from the same amount of hardware we're using.

Lev: For now, I think the immediate focus needs to be on proving that this protocol works with enough fidelity on noisy chips before we worry about pushing past the four-unit limit.

Mira: Well, "Superconducting-antiqubit simulation enables optimal phase estimation via unitary inversion" definitely shows how entanglement and inversion combine for unique sensing power.

Kai: That’s what we’ve got today on quantum sensing with simulated antimatter. Next up, we'll be looking at some instability issues in self-consistent diagrammatic approaches.

Department of Physics, Washington University, St. Louis, Missouri 63130, USA · Cavendish Laboratory, Department of Physics, University of Cambridge, Cambridge CB3 0HE, United Kingdom · Joint Center for Quantum Information and Computer Science NIST and University of Maryland · Institute for Physical Science and Technology University of Maryland

quant-ph, cond-mat.mes-hall, hep-ex, physics.atom-ph

Submitted: 2025-06-04

Updated: 2026-10-08

Comments: 34 pages, 10 figures

Code: https://github.com/FacebookResearch/Nevergrad

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Importance score: 89/100

The gist: The gist: Positronium metrology achieves an FI of 4 while consuming vst = 2 units of space–time volume, which equals the greatest possible QFI achievable with any strategy, regardless of whether

Key concepts

Antiqubits: These are simulated transmon qubits engineered to behave like positrons. They achieve this by using specific resonant and off-resonant drives that result in an effective gyromagnetic ratio equal to the negative of a standard qubit's ratio. This allows them to simulate antimatter properties within a superconducting circuit.
Platform-Specific Unitary Inversion: This is a method where the environment (the drive fields) performs a time inversion on the unitary operation applied by an external magnetic field. By using antiqubits, this inversion can be achieved via entanglement with the qubit, effectively inverting the action of $U_eta$ to $U^eta$.
Fisher Information (FI): FI is a measure of how much information a quantum measurement can extract about an unknown parameter, like the direction of a field. The paper shows that using entangled qubits and antiqubits allows for the highest possible FI per unit of space-time volume, reaching a theoretical maximum of 4.
Positronium Metrology: This is the specific sensing strategy implemented using these simulated antimatter systems. It involves preparing a singlet state, applying specific unitary operations to the qubit and antiqubit, and measuring the resulting probability. This process yields an optimal sensitivity for estimating unknown field parameters.

Terminology

Entanglement — 1× in this paper · explained in 4 episodes
Fisher Information — 1× in this paper · explained in 3 episodes

Summary

The gist: Positronium metrology achieves an FI of 4 while consuming vst = 2 units of space–time volume, which equals the greatest possible QFI achievable with any strategy, regardless of whether the field direction is known<ref:2506.04315#pg11>

Platform-Specific Unitary Inversion

The paper introduces antiqubits, which are simulated transmon superconducting qubits subject to particular resonant and off-resonant drives, where an antiqubit’s effective gyromagnetic ratio equals the negative of a qubit’s<ref:2506.04315#pg4> This fact enables the time inversion of a unitary implemented on a transmon by its environment<ref:2506.04315#pg5> The strategy relies on resonant and offresonant drives, which negate the effective gyromagnetic ratio of the transmon’s pseudospin, thereby inverting the unitary to which a magnetic field subjects the transmon<ref:2506.04315#pg5> This process is called platform-specific unitary inversion, and these systems are termed positronium-like transmons or antiqubits<ref:2506.04315#pg6>

Positronium Metrology and Quantum Advantage

The core finding demonstrates that entanglement and platform-specific unitary inversion enable a quantum advantage in sensing, specifically in phase estimation<ref:2506.04315#pg7> By entangling a qubit transmon with an antiqubit, the field applies Uα to the qubit and, via platform-specific unitary inversion, U†α to the antiqubit<ref:2506.04315#pg6> This boosts the Fisher information (FI) obtainable or sensitivity achievable about the field<ref:2506.04315#pg7> The strategy achieves an FI of 3.02 per two units of space–time volume experimentally, which is greater than that achievable by two entangled qubits (1) or a separable qubit–antiqubit state (4/3)<ref:2506.04315#pg7>

Theoretical Framework for Optimal Sensing

The optimal FI per two units of space-time volume is achieved when entanglement and effective unitary inversion are combined<ref:2506.04315#pg7> The theoretical limit on this metric is 4, which saturates the bound achievable with any strategy consuming ≤ 2 units of space–time volume, even if one knows nˆ<ref:2506.04315#pg7> This maximum QFI is achieved when the input state is a singlet and one TLS undergoes Uα while the other undergoes U†α<ref:2506.04315#pg18>

Optimal Input States and Strategy

The optimal two-TLS input states consistent with a fixed concurrence C0 are characterized by specific relative rotations Rrel that maximize the QFI<ref:2506.04315#pg16> For C0 < 1, the QFI is maximized when nˆ = yˆ (if ss′ = −1) or nˆ = xˆ (if ss′ = 1), leading to an FI of I(s)α = 2(1+C0)<ref:2506.04315#pg16> When C0 = 1, the maximal QFI is determined entirely by the contribution proportional to sC0, achieving I(s)α = 4 regardless of nˆ<ref:2506.04315#pg16> Positronium metrology exemplifies one of these optimal-QFI conditions when C0 = 1 and s = −1, resulting in the singlet state

Experimental Realization

The experimental setup involves three transmons: a qubit q acting as ordinary matter, an antiqubit ¯q, and a coupler c<ref:2506.04315#pg5> The singlet state is prepared by subjecting q to a π-pulse, followed by a √iSWAP gate via resonant parametric interaction for 104 ns, and then correcting the antiqubit’s phase<ref:2506.04315#pg6> Platform-specific unitary inversion is implemented using Z gates and off-resonant microwave drives to invert magnetic field components The magic frequency of the drive induces equal-magnitude, opposite-sign AC Stark shifts on q and ¯q, allowing for the simulation of antimatter<ref:2506.04315#pg22>

Conclusion and Outlook

The work establishes that effective unitary inversion and entanglement enable the unique optimal sensing strategy available when nˆ is unknown and vst = 2<ref:2506.04315#pg18> This strategy's QFI equals the greatest possible QFI, Iα = 4<ref:2506.04315#pg18> Future research will explore extending this protocol to achieve more FI per unit space–time volume and sketching uses of actual antimatter in quantum sensing This work also suggests applications for unitary-inversion methods beyond metrology, including black-box access algorithms and cryptographic scheme breaking<ref:2506.04315#pg16>

How it works

The first effective time reversal results from entanglement manipulation and enables agnostic phase estimation by effectively teleporting the probe state backward in time via entanglement manipulations<ref:2506.04315#pg10> The second effective time reversal is achieved by converting the ancilla into an antiqubit by negating its effective gyromagnetic ratio, subjecting it to U†α<ref:2506.04315#pg11> Implementing positronium metrology is equivalent to implementing U2α on the probe, which costs one unit of space–time volume instead of two units for true U2α<ref:2506.04315#pg18>

QFI Achievable via Positronium Metrology

Positronium metrology involves three steps: preparing a synthetic positronium atom in a singlet, applying the unitary Uα to q and U†α to ¯q, and measuring the POVM<ref:2506.04315#pg22> The probability of obtaining the outcome associated with the first projector is P(Ψ−⟩) = cos2(α)<ref:2506.04315#pg19> Substituting this into the FI formula yields Iα = 4 cos2(α) sin2(α) / cos2 (α) sin2(α) = 4<ref:2506.04315#pg19>

QFI Achievable with Synthetic Antimatter and No Entanglement

The entanglement-free protocol, using a product state x+⟩q ⊗ z+⟩¯q, yields an effective QFI of Iα,eff = 5/6<ref:2506.04315#pg21> This is slightly lower than the optimal QFI achievable without entanglement but with a space–time volume vst = 2, which is 4/3 ≈ 1.33<ref:2506.04315#pg20> Positronium metrology achieves a QFIM of the form (81) and Iα = 4<ref:2506.04315#pg20>

Experimental Realization of Qubits and Antiqubits

The qubit transmon frequency is fixed, while c and ¯q are flux-tunable Dispersive readout allows for simultaneous high-fidelity single-shot readouts with fidelities of 97.8 % for the qubit and 95.0 % for the antiqubit<ref:2506.04315#pg18> Single-qubit rotations are implemented using various pulse envelopes, including π/2 rotations about x- and y-axes<ref:2506.04315#pg19> Parametric gates induce qubit–antiqubit coupling at a rate of 54.0 MHz when the coupler is modulated<ref:2506.04315#pg22>

Data Availability and Author Contributions

The data that support the findings of this study are available from the corresponding authors upon reasonable request<ref:2506.04315#pg8> The authors contributed to the writing and editing of the manuscript, with F.S. crafting the theory sections and proofs<ref:2506.04315#pg9> The work received support from various grants including QLCI grant OMA-2120757<ref:2506.04315#pg16>

Competing Interests

The authors declare no competing interests<ref:2506.04315#pg10> This research was supported in part by grant NSF PHY-2309135 to the Kavli Institute for Theoretical Physics (KITP)<ref:2506.04315#pg9> The work was supported by the National Science Foundation through QLCI grant OMA-2120757, PHY-2408932, and by the Gordon and Betty Moore Foundation<ref:2506.04315#pg16> The authors acknowledge support from a QuICS Hartree Postdoctoral Fellowship<ref:2506.04315#pg9> The work was supported in part by grant NSF PHY-2309135 to the Kavli Institute for Theoretical Physics (KITP)<ref:2506.04315#pg9> The work was supported by the National Science Foundation through QLCI grant OMA-2120757, PHY-2408932, and by the Gordon and Betty Moore Foundation<ref:2506.04315#pg16> The authors acknowledge support from a QuICS Hartree Postdoctoral Fellowship<ref:2506.04315#pg9> The work was supported in part by grant NSF PHY-2309135 to the Kavli Institute for Theoretical Physics (KITP)<ref:2506.04315#pg9>

References

[1] E. C. C. St¨uckelberg. La Mecanique du point materiel en theorie de relativite et en theorie des quanta, Helv. Phys. Acta, 15:23–37, 1942<ref:2506.04315#pg10>

[2] R. P. Feynman. A relativistic cut-Off for classical electrodynamics, Phys. Rev., 74:939–946, 1948<ref:2506.04315#pg11>

[3] R. P. Feynman. The theory of positrons, Phys. Rev.

Improvements for AI systems

textbfAmplified Phase Estimation Sensitivity via Positronium Metrology (FI=4): The improved AI system can achieve a greatest possible QFI for phase estimation, specifically achieving an FI of 4 while consuming only two units of space-time volume, which is the theoretical maximum achievable for two units of space-time volume.
textbfDirection-Independent Quantum Sensing: Robust Phase Estimation without Field Knowledge: The system can perform quantum sensing even when the direction of an external field is unknown by leveraging entanglement and platform-specific unitary inversion, which allows it to achieve an FI of 4 regardless of the rotation axis, unlike entanglement-free strategies which might only achieve 1.3 ± 0.1.
textbfhigher-order Quantum Transformation Implementation: Advanced Quantum Channel Simulation: The system can implement higher-order quantum transformations (e.g., evolutions of quantum channels) by effectively inverting unitaries, enabling the simulation of complex time evolution and potentially supporting quantum singularvalue transformations.