Operational Characterizations and Resource Structure of the Interactive Weight of Quantum Instruments

Technical Summary: Operational Characterizations and Resource Structure of the Interactive Weight of Quantum Instruments

Problem Statement
Quantum instruments describe measurement processes that yield both classical outcomes and quantum state transformations. While the resource theory of interactive quantum instruments has been established based on the perspective of robustness (quantifying the noise required to render an instrument non-interactive), a systematic characterization based on weight remains underexplored. Specifically, there is a need to quantify the maximal fraction of a quantum instrument that can be explained by non-interactive components (free resources) and to understand the operational significance and compositional properties of this "non-interactive content." Existing frameworks have not fully addressed the exact operational interpretation of this complementary quantity or its behavior under multicopy resource conversion.

Methodology
The authors introduce the interactive weight ($W_{\text{int}}$) of a quantum instrument as a resource quantifier defined via convex decomposition. An instrument $\mathcal{E}$ is decomposed as $\mathcal{E} = (1-s)\mathcal{L} + s\mathcal{G}$, where $\mathcal{L}$ is a non-interactive instrument (the free set) and $\mathcal{G}$ is an arbitrary instrument. $W_{\text{int}}(\mathcal{E})$ is the minimum $s$, while its complement, $\lambda_{\text{NI}}(\mathcal{E}) = 1 - W_{\text{int}}(\mathcal{E})$, represents the maximal non-interactive content.

Key methodological steps include:

1. Semidefinite Programming (SDP) Formulation: Leveraging the specific product structure of the Choi operators for non-interactive instruments ($J_{\mathcal{L}}^a = \frac{I_{A'}}{d_A} \otimes \tau_B^a$), the authors derive primal and dual SDP formulations for $W_{\text{int}}$. Strong duality is established, ensuring computability.
2. Operational Scenarios: The paper investigates four distinct operational tasks to characterize the resource:
   • Maximally Entangled-State Recovery: Quantifying the fidelity of recovering a maximally entangled state after the instrument acts on one half.
   • Haar-Averaged Pure-State Recovery: Quantifying the average fidelity of recovering unknown pure states.
   • Positive Linear Loss Games: A prepare-instrument-measure game where the loss is evaluated against a non-interactive benchmark.
   • Environment-Assisted Input-Blind Extraction: A scenario where an assistant measures the minimal Stinespring environment to postselect a subprocess that is completely independent of the input state (fully input-blind).
3. Resource Theory Framework: The authors analyze the behavior of $W_{\text{int}}$ under the class of deterministic allowed operations (AO) introduced by Hsieh et al., which include preprocessing, classical outcome processing, and conditional postprocessing.

Key Contributions and Results

• Structural Properties:

  • Faithfulness and Convexity: $W_{\text{int}}$ is faithful (zero if and only if the instrument is non-interactive) and convex under probabilistic mixtures.
  • Exact Multiplicativity: The maximal non-interactive content satisfies an exact multiplicative law under independent parallel composition: $\lambda_{\text{NI}}(\mathcal{E} \otimes \mathcal{F}) = \lambda_{\text{NI}}(\mathcal{E})\lambda_{\text{NI}}(\mathcal{F})$. This contrasts with robustness-based measures which often satisfy only inequalities.

• Operational Characterizations:

  • Recovery Bounds: Quantitative bounds are established linking $W_{\text{int}}$ to recovery fidelities in both entangled-state and Haar-averaged pure-state tasks. These bounds yield a unified quantitative relation between interactive weight and interactive instrument robustness ($R_{\text{int}}$): $W_{\text{int}}(\mathcal{E}) \geq \frac{R_{\text{int}}(\mathcal{E})}{d_A^2 - 1}$.
  • Exact Characterization via Loss Games: The maximal non-interactive content is exactly characterized as the infimum of the normalized loss ratio of the target instrument relative to the non-interactive benchmark over all admissible positive linear loss games.
  • Environment-Assisted Extraction: The paper proves that $\lambda_{\text{NI}}(\mathcal{E})$ is exactly equal to the maximum success probability of extracting a fully input-blind subprocess from the instrument via environmental postselection on the minimal Stinespring environment. This provides a direct probabilistic interpretation of the resource.
  • Connection to Doeblin Coefficient: The maximal non-interactive content is shown to be equivalent to the positive quantum Doeblin coefficient ($\alpha_+$) of the associated classical-quantum channel retaining the instrument's native classical outcome.
  • Restricted Access: The framework is extended to scenarios with restricted environmental access, yielding a universal upper bound on extraction probability and an access data-processing inequality.

• Resource Conversion and Distillation:

  • Monotonicity: $W_{\text{int}}$ is monotonic under deterministic allowed operations, while $\lambda_{\text{NI}}$ is non-decreasing.
  • Additive Measure: An additive logarithmic interactivity measure, $I_{\text{int}}(\mathcal{E}) = -\log_2 \lambda_{\text{NI}}(\mathcal{E})$, is defined.
  • Multicopy Constraints: The exact multiplicativity of $\lambda_{\text{NI}}$ imposes necessary constraints on exact multicopy resource conversion. Specifically, converting $n$ copies of an instrument $\mathcal{E}$ to $m$ copies of $\mathcal{N}$ requires $n I_{\text{int}}(\mathcal{E}) \geq m I_{\text{int}}(\mathcal{N})$.
  • No-Go Theorem: The authors establish a finite-copy maximal-weight distillation no-go result. Under deterministic allowed operations, a finite number of copies of an instrument with non-maximal interactive weight ($W_{\text{int}} < 1$) cannot be exactly converted into an instrument with maximal interactive weight ($W_{\text{int}} = 1$). This is because the non-interactive content of the source remains strictly positive for any finite number of copies, preventing exact distillation to a state with zero non-interactive content.

Significance
The paper establishes a unified theoretical framework for the interactive weight of quantum instruments, complementing the existing robustness-based approach. By defining a weight-based quantifier, the authors provide:

1. Operational Clarity: A direct probabilistic interpretation via environment-assisted extraction and exact characterizations via loss games.
2. Compositional Insight: The discovery of exact multiplicativity for the complementary non-interactive content, which enables the definition of an additive resource measure.
3. Fundamental Limits: The derivation of strict constraints on multicopy resource conversion, specifically proving the impossibility of exact finite-copy distillation to maximal-weight instruments under deterministic operations.

These results offer systematic theoretical tools for the quantitative comparison and conversion of interactive quantum instruments, bridging the gap between convex decomposition structures and operational tasks. The authors note that while the current framework focuses on exact extraction in single-use or finite-copy settings, the extension to approximate and asymptotic settings remains an open area for future investigation.