Simulation of Adjoints and Petz Recovery Maps for Unknown Quantum Channels

Technical Summary: Simulation of Adjoints and Petz Recovery Maps for Unknown Quantum Channels

Problem Statement
The paper addresses a fundamental problem in quantum information theory: the physical realizability of transformations on unknown quantum channels, specifically the complex conjugate ($N^*$), transpose ($N^T$), and adjoint ($N^\dagger$). While these transformations are well-understood for unitary operations (where $U \to U^\dagger$ corresponds to time reversal), their implementation for general open-system dynamics (quantum channels) presents distinct challenges. A primary difficulty is that while the complex conjugate of a channel remains a completely positive (CP) and trace-preserving (TP) map, the transpose and adjoint are generally not TP, meaning they do not constitute valid quantum channels. The central question is whether these non-TP maps can be implemented physically via probabilistic supermaps or if they require virtual (quasi-probabilistic) schemes.

Methodology
The authors employ a hierarchy of higher-order quantum transformations, distinguishing between:

1. Probabilistic Combs: Physical protocols that succeed with a certain probability (post-selection).
2. Virtual Combs: Protocols based on quasi-probability decompositions, where physical processes are combined with negative weights to simulate non-physical maps.

The methodology involves:

• Post-selected Teleportation: Utilizing maximally entangled states and Bell measurements to probabilistically implement the transpose of an unknown channel.
• No-Go Proofs: Using proof by contradiction to demonstrate that no completely positive (CP) supermap, even with finite copies of the input channel, can universally realize the complex conjugate or adjoint.
• Werner-Holevo Channels: Constructing a virtual comb using Werner-Holevo channels ($W^\pm_d$) as pre- and post-processing steps. These channels mix the desired transpose term with isotropic noise. By taking a linear combination (quasi-probability mixture) of these channels, the noise terms are cancelled, isolating the non-physical complex conjugate map.
• Diamond Norm Optimization: Quantifying the deviation of the virtual comb from physical combs using the base norm (equivalent to the diamond norm for combs) to prove the optimality of the proposed construction.
• Petz Recovery Estimation: Integrating the virtual adjoint protocol with block-encoding techniques to estimate expectation values of the Petz recovery map without requiring access to the Stinespring isometry of the channel.

Key Contributions and Results

1. Strict Hierarchy of Realizability:

   • Transpose ($N^T$): Proven to be physically realizable via a probabilistic post-selected teleportation protocol. The success probability is $1/d^2$ for unital channels.
   • Complex Conjugate ($N^*$) and Adjoint ($N^\dagger$): The paper proves a "no-go" theorem (Theorem 1) stating that neither $N^*$ nor $N^\dagger$ can be implemented by any completely positive supermap, even probabilistically, for an unknown channel. This establishes a fundamental gap between unitary and general channel transformations.

2. Virtual Protocol for Complex Conjugation:

   • The authors design a 1-slot virtual comb (Theorem 3) that universally implements the complex conjugate $N^*$ using a quasi-probability decomposition of Werner-Holevo channels.
   • The protocol involves sampling from three physical process combinations with specific probabilities ($p_1, p_2, p_3$) and classical post-processing to reconstruct the expectation value $\text{Tr}[O N^*(\rho)]$.
   • Optimality: Theorem 4 establishes that the constructed virtual comb achieves the minimum possible base norm (sampling overhead) of $d_A d_B - d_A + 1$, proving the construction is optimal with respect to the diamond norm.

3. Petz Recovery Map Estimation:

   • The paper proposes a protocol to estimate the expectation value $\text{Tr}[O_A P_{\sigma, N}(\omega_B)]$ for the Petz recovery map of an unknown channel $N$, given a prior state $\sigma$.
   • Unlike previous methods (e.g., Ref. [30, 31]) that rely on block-encoding the Stinespring isometry or deterministic approximations with high query complexity, this method uses only black-box access to the channel.
   • Query Complexity: For unital channels, the query complexity scales as $O\left(\frac{d_A^3 d_B^3}{\varepsilon^2} \log \frac{1}{\delta}\right)$. This represents a significant improvement over the deterministic approximation in Ref. [31], which scales as $O\left(\frac{d_A^{5.5} d_B^{2.5}}{\varepsilon^4 \lambda_{\min}^{3/2}} \dots\right)$.

Significance
The paper claims to provide a complete picture of the physical implementation of universal dual maps (conjugate, transpose, adjoint) for quantum channels. Its significance lies in:

• Resolving Fundamental Limits: It definitively separates the capabilities of physical supermaps (which can handle the transpose but not the conjugate/adjoint) from virtual protocols (which can handle all three).
• Enabling New Applications: By providing a virtual protocol for the adjoint, the work enables the estimation of the Petz recovery map using only black-box channel access. This is crucial for characterizing the reversibility of quantum dynamics and near-optimal error correction.
• Efficiency: The proposed method offers a polynomial improvement in query complexity for estimating Petz map expectation values compared to existing deterministic approaches, making it more feasible for practical quantum information processing tasks involving open systems.

The authors note that this framework allows for the probing of out-of-time-ordered correlators (OTOCs) in generic open quantum systems and suggests future extensions to multi-slot settings and broader classes of higher-order transformations in quantum thermodynamics and error mitigation.