arXiv:2610.20528
Quantum Work Statistics from Entanglement-Assisted Invariance
Akira Sone, Akram Touil, Emery Doucet, Sebastian Deffner
Abstract
We demonstrate that the one-time measurement scheme is more consistent than the standard two-time measurement scheme with respect to the fundamental symmetry of entanglement, namely entanglement-assisted invariance (envariance). In particular, the work statistics obtained from the one-time measurement scheme arise naturally as a direct consequence of envariance for any driving unitary, without involving an explicit measurement postulate, whereas the analogous identification for the two-time measurement scheme is possible only for a Haar-measure-zero set of unitaries. This envariance-based formulation eliminates the ambiguity in work statistics related to the energetic cost of measurement.
AI Summary
The paper asks whether entanglement-assisted invariance (envariance), used to motivate Born probabilities, distinguishes one-time measurement (OTM) and two-time measurement (TTM) definitions of quantum work. It compares their coherent implementations using a finite-dimensional driven system coupled to a work meter whose displacement records work. For OTM, the final coupling is controlled in the evolved initial-energy basis and records the conditional energy change $\widetilde w_n=\langle n|U^\dagger H_\tau U|n\rangle-E_n$. The resulting state is already in Schmidt form after grouping equal work values, with squared Schmidt coefficients equal to the OTM probabilities for every driving unitary. The construction recovers the mean energy change and the previously established OTM exponential-work relation. For the TTM coupling, the system states associated with different meter readings are generally nonorthogonal. If the initial state populates at least two distinct energies and the final Hamiltonian has at least two distinct levels, mutual orthogonality occurs only on a Haar-measure-zero subset of $\mathrm{U}(d)$. This follows by expressing branch overlaps as analytic functions and constructing a unitary with nonzero overlap. The distinction is specific to the chosen system–meter construction: adding a reference that retains the initial-energy label also puts TTM statistics into Schmidt form. The associated additional information cost is left for future work. The treatment uses idealized orthogonal meter readings with a discretization approximation; it does not provide a quantitative energetic-cost comparison of the two implementations.
Central claim may not hold: The measure-zero result applies only to the chosen two-part meter construction, while the paper's own reference-system construction gives TTM probabilities as squared Schmidt coefficients for arbitrary driving unitaries. The abstract omits essential spectral and initial-support assumptions, without which TTM branch orthogonality can hold for every driving unitary. The claimed elimination of measurement-energy-cost ambiguity is unsupported because the controlled meter interactions are not costed and the reference-system cost analysis is deferred. For degenerate initial energies, the initial meter coupling preserves within-energy coherences that the paper incorrectly discards, invalidating its diagonal reduced-state and TTM-weight formulas in general.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 4.5 + 4 pages