Hidden Fisher Information in Many-Body Quantum Sensors

AI summary

Overview

Research area: Quantum metrology and quantum sensing, specifically many-body collective-spin sensors, quantum Fisher information (QFI), and measurement/readout complexity.

Technical level: Advanced. The paper assumes familiarity with the quantum Cramér–Rao bound, the symmetric logarithmic derivative (SLD), SU(2) collective spin algebra, and irreducible tensor operators.

Scope (one sentence): The paper introduces a degree-resolved "moment Fisher hierarchy" that quantifies how much of a sensor's QFI is actually accessible to collective spin measurements of bounded degree, and uses a kicked-top model to show that nonlinear many-body dynamics can hide large QFI in high-rank operator sectors.

What This Paper Is About

The QFI sets the best precision any quantum sensor can reach, but it assumes a completely unconstrained, fully optimized measurement, so it says nothing about how hard that optimal measurement is to build. In many-body systems the optimal readout (the SLD) can be a highly nonlocal, high-rank operator that barely overlaps with what experiments actually measure, such as mean spin, spin noise, or low-order correlations. The authors ask when a sensor's missing sensitivity is a state-preparation problem versus a readout problem, and they build a quantitative diagnostic to separate the two.

Key Contributions

1. A moment Fisher hierarchy. They define $F_M^{(k)}$, the maximum error-propagation Fisher information obtainable from any scalar estimator built from collective spin polynomials of degree at most $k$, and the normalized accessibility $\eta_k = F_M^{(k)}/F_Q$, which measures the fraction of the QFI visible to moments up to degree $k$.
2. An operational readout-complexity scale. They introduce $k_{90} = \min{k : F_M^{(k)} \geq 0.9 F_Q}$, the minimum moment degree needed to recover 90% of the QFI — a single number characterizing how demanding the readout is, analogous to a resolution requirement.
3. A demonstration that many-body dynamics can hide QFI. In a kicked collective spin (a symmetric-subspace quantum kicked top), they show that a resonant nonlinear protocol preserves large QFI while pushing it out of reach of low-order collective observables, up to $N = 100$ spins.
4. A mechanistic and basis-independent explanation. They identify "nonlinear tangent-space shear" as the cause, and confirm it with a rank-resolved decomposition of the SLD into irreducible SU(2) tensor sectors under weak damping, showing the optimal readout migrates to the highest available ranks.

Main Findings

• Linear protocol is readout-simple. For the linear baseline $\beta = 0$, first moments nearly saturate the QFI and $k_{90} = 1$, so $k_{90}/N$ decreases with system size — the expected behavior when the absolute readout degree stays fixed.
• Resonant nonlinear protocol hides the information. For $\beta = \pi j$, first moments access only about $10^{-2}$ of the QFI, reaching $\eta_1 \simeq 10^{-2}$ at $N = 100$, even though the QFI itself grows strongly with $N$.
• Readout depth grows with system size. At $N = 100$ the data give $k_{90} \simeq 28$, i.e. $k_{90}/N \simeq 0.28$. The authors explicitly state these finite-size data are not used to assume a specific asymptotic scaling law.
• The information is not lost, only redistributed. The endpoint of the hierarchy satisfies $F_M^{(N)} = F_Q$ within numerical precision (verified for every hierarchy curve in the figures), while $F_M^{(1)}$ and $F_M^{(2)}$ stay far below $F_Q$; the gap is therefore inaccessible readout structure, not numerical loss of QFI.
• The optimal readout moves to high tensor rank. In the rank-resolved SLD diagnostic with weak collective damping ($\gamma = 10^{-3}$, $N \leq 20$), the nonlinear protocol gives $\ell_{90}/N = 1$ over $N = 2$–$20$, meaning 90% of the SLD weight requires the maximal available rank, whereas the linear protocol reaches the same cumulative weight at much smaller normalized rank.
• Low-rank SLD weight is strongly depleted. For $\beta = \pi j$, $W_{\leq 2}$ is suppressed from order unity at the smallest system to the $10^{-3}$ scale over most of the studied range, while the linear baseline retains order-unity low-rank weight.
• The barrier is not an artifact of exact resonance. A detuning scan over $r = \beta/(\pi j)$ shows that once nonlinear kicking is introduced, $\eta_1$ is strongly suppressed; at $N = 100$ it lies at the percent level or below for much of the nonlinear regime, with the strongest barrier near $r \simeq 1$.
• The hierarchy is a resource diagnostic, not a no-go theorem. The authors stress that number-resolved histograms, parity measurements, tomography, randomized measurements, or interaction-based decoding can access information beyond $F_M^{(k)}$ at fixed low $k$; the Supplemental Material shows such collective readouts recover substantially more than linear moments.

Methodology in Plain English

The authors study $N$ spin-1/2 particles restricted to the permutation-symmetric manifold with total spin $j = N/2$. One Floquet step is a parameter rotation $e^{-i\alpha J_z}$ followed by a one-axis-twisting kick $e^{-i\beta J_y^2/(2j)}$, starting from a spin coherent state with $\Theta = \pi/3$, $\Phi = 0$, evaluated at $\alpha = \pi/2$, with eight kicks per cycle; the linear case $\beta = 0$ is compared with the resonant case $\beta = \pi j$.

To build the hierarchy, they enumerate all symmetrized collective spin products up to degree $k$, compute how strongly the state's parameter derivative responds to each one, and compute the noise (covariance) between them. Maximizing the signal-to-noise ratio over all linear combinations gives $F_M^{(k)}$ as a matrix expression involving the response vector and the pseudoinverse of the covariance matrix — effectively asking "what is the best scalar built from low-degree collective moments?" Because moments of degree $N$ span the entire traceless operator space (dimension $N(N+2)$, in a Hilbert space of dimension $N+1$), the hierarchy must recover the full QFI at $k = N$, which provides a built-in consistency check. For large $N$ they use a pure-state version that projects the state's tangent vector onto the span generated by degree-$k$ observables, working in an SU(2) tensor basis to avoid the severe ill-conditioning of high-degree monomials in $J_x, J_y, J_z$. As an independent check, they add weak collective damping ($\gamma = 10^{-3}$) for $N \leq 20$ and decompose the SLD into irreducible tensor ranks.

Why This Matters

Impact on research. The QFI has been the standard benchmark for quantum-enhanced metrology, but it silently averages over the cost of implementing the optimal measurement. This work makes readout complexity a measurable quantity, giving theorists and experimentalists a shared language for the question "is my sensor limited by the state or by the detection?" It also reinterprets existing advanced-readout techniques — interaction-based readout, phase magnification, time reversal, parity detection, number resolution — as either physically rotating a high-rank SLD back into a simple observable or measuring finer structure directly, closing the gap between $F_Q$ and $F_M^{(1)}$ or $F_M^{(2)}$.

Real-world applications (platforms discussed in the paper):

• Optical clocks and Rydberg arrays, where strong dipole interactions generate states with large metrological capability but readout is typically built around a few collective observables.
• Cavity-QED ensembles, where light-matter interactions create robust entangled states with programmable interactions.
• Trapped ions and spinor condensates, where metrologically useful correlations have been demonstrated and final sensitivity often hinges on the readout stage.
• Non-Gaussian, interaction-generated probes more generally, where fine angular structure on the collective Bloch sphere is invisible to spin-squeezing or covariance measurements.

Industry relevance. For quantum sensing and timing companies, the hierarchy offers a practical diagnostic for deciding whether to invest in better state preparation or in better detection hardware: if $F_M^{(1)}$ or $F_M^{(2)}$ nearly saturates $F_Q$, standard collective detection suffices; if $F_M^{(k)}$ stays far below $F_Q$ until $k$ is large, the apparatus needs more readout resolution. The paper also warns that a null result in conventional spin-noise measurements should not automatically be read as absence of metrological value.

Future Directions

1. Experimental test of hidden Fisher information. The authors propose measuring the same state before and after an echo, time-reversal, or interaction-based decoding sequence; closing the hierarchy gap would directly demonstrate conversion of high-rank metrological information into a readable low-degree signal.
2. Asymptotic scaling of $k_{90}$. The paper establishes growth of readout depth over $N \leq 100$ but explicitly declines to assume a scaling law; determining whether $k_{90}$ scales linearly or otherwise with $N$ remains open.
3. Interplay with decoherence. Dissipation acts as a low-pass filter that suppresses high-rank tensor components faster than low-rank ones, so at fixed nonzero damping ultra-high-rank information may eventually degrade; where the crossover lies, and how it depends on $\gamma$, is not resolved.
4. Generalization beyond the kicked top. Because the mechanism relies on nonlinear generation of high-rank collective operator structure, the authors expect it to arise broadly in non-Gaussian many-body sensors; testing the hierarchy in other driven or interaction-dominated platforms is a natural next step.

Target Audience

Quantum metrology researchers and theorists working on many-body sensing, spin squeezing, and non-Gaussian probe states; experimentalists in cold atoms, trapped ions, cavity QED, Rydberg arrays, and optical clocks who need to decide whether their sensitivity is preparation-limited or readout-limited; and graduate students or advanced readers with a background in quantum estimation theory and collective spin algebra who want a quantitative handle on measurement complexity in many-body quantum sensors.

Authors’ abstract

Quantum Fisher information (QFI) determines the ultimate precision of a quantum sensor after optimization over all allowed measurements, but it does not quantify how difficult the optimal readout is to implement. We introduce a moment Fisher hierarchy $F_M^{(k)}$, which measures the fraction of QFI accessible to collective spin moments up to degree $k$, together with a readout-complexity scale $k_{90}$ defined by the minimum degree required to recover $90\%$ of the QFI. Applying this framework to a nonlinear collective-spin sensor, we show that many-body dynamics can preserve large QFI while transferring the optimal readout into high-rank $SU(2)$ tensor sectors inaccessible to low-order collective observables. For system sizes up to $N=100$, first moments access only $\sim 10^{-2}$ of the available QFI, while the degree required to recover $90\%$ of it grows to $k_{90}\simeq 28$. A rank-resolved analysis of the symmetric logarithmic derivative (SLD) under weak damping shows that the nonlinear protocol pushes the optimal readout toward the highest available tensor ranks while strongly depleting its low-rank weight. These results identify nonlinear tangent-space shear as the mechanism behind a finite-size readout-complexity bottleneck in many-body quantum metrology.