Advantages of the Kirkwood–Dirac distribution among general quasi-probabilities on finite-state quantum systems

Shun Umekawa, Jaeha Lee, Naomichi Hatano

TL;DR

This paper argues that the Kirkwood-Dirac (KD) distribution is a superior quasi-probability distribution for finite-state quantum systems compared to others like the Wigner function. It demonstrates that the KD distribution behaves more like a true probability distribution and provides richer information, particularly for distinguishing quantum states.

What Problem Does This Paper Address?

In quantum mechanics, a true joint probability distribution for multiple non-commuting observables cannot be defined. To gain intuition, physicists use "quasi-probability distributions" (like the Wigner function). However, these distributions can take on unphysical values (e.g., negative probabilities) and their behavior varies significantly depending on the system and the specific quasi-probability chosen. This paper addresses the question of which quasi-probability distribution is most "well-behaved" and informative, especially for finite-state quantum systems (like qubits or systems with a limited number of energy levels). The authors specifically investigate if there's a preferred quasi-probability that mimics classical probability more closely and offers better state distinguishability.

Methods

The authors work within the general framework of "quasi-classicalization," which unifies various quasi-probability distributions. They analyze different types of quasi-joint-probability distributions (QJSDs) by examining their properties, particularly for two-state (spin-1/2) and three-state quantum systems. Key aspects of their analysis include:

  • Comparing QJSDs: They compare the Kirkwood-Dirac distribution with others, including the Wigner function, by looking at their behavior for specific quantum systems.
  • "Nice" Behavior: They define "nice" behavior as resembling a genuine probability distribution (e.g., taking non-zero values only at physically allowed outcomes) and providing sufficient information to distinguish quantum states.
  • State Distinguishability: They investigate how many observables are needed to uniquely identify a quantum state using a given QJSD.
  • Real vs. Complex Values: They analyze when QJSDs yield real values, linking this to the physical properties of the quantum state.

Datasets

Not discussed in the paper. The paper focuses on theoretical analysis of quantum systems rather than empirical data.

Key Insights and Ideas

  • Kirkwood-Dirac's Advantage: The KD distribution consistently shows "nicer" behavior for finite-state systems compared to other QJSDs. It takes non-zero values only at the physically allowed outcomes of observables, much like a classical probability distribution.
  • State Distinguishability: For two- and three-state systems, the KD distribution of just two spin components is sufficient to distinguish all possible states. This is a significant advantage over methods relying solely on expectation values, which might require more observables.
  • Imaginary Part Matters: For two-state systems, the imaginary part of the KD distribution provides crucial information for distinguishing states, highlighting its non-classical nature.
  • Real QJSDs and State Properties: The condition for a KD distribution to be real is directly linked to specific properties of the quantum state (e.g., lying in the "xy-plane" of the Bloch sphere for spin-1/2 systems).
  • Generalizability: The properties that make the KD distribution advantageous for two- and three-state systems appear to generalize to arbitrary finite-state quantum systems.

Key Results

  • For two- and three-state quantum systems, the Kirkwood-Dirac distribution is shown to be superior to other quasi-probability distributions in terms of mimicking classical probability and providing state distinguishability.
  • The states of two- and three-state quantum systems can be uniquely distinguished using the Kirkwood-Dirac distribution of only two spin components.
  • For two-state systems, the imaginary part of the Kirkwood-Dirac distribution is essential for distinguishing states.
  • The paper establishes conditions under which quasi-joint-probability distributions are real-valued, linking these to the properties of the quantum state.

Limitations and Future Work

  • Higher-Dimensional Systems: While the paper suggests that the advantages of the KD distribution might generalize to higher-dimensional Hilbert spaces, this is not rigorously proven for all cases. The authors pose this as an open question.
  • Infinite-State Systems: The relationship of these findings to quasi-probabilities for infinite-state systems (like those in optical systems) is also left for future investigation.
  • Specific Observable Pairs: While the paper shows that any pair of non-commutative observables can be used for state distinguishability with the KD distribution in spin-1 systems, finding specific pairs that don't belong to the spin-1 representation of su(2) but still work is an open area.

Summary updated

Abstract

Abstract We investigate the properties of quasi-joint-probability (QJP) distributions on finite-state quantum systems, especially two- and three-state systems, based on the general framework of quantum/quasi-classical representations. We show that the Kirkwood–Dirac distribution is a prime candidate among the QJP distributions that behave well in view of the following two perspectives: the information contained in the QJP distribution and its affinity to genuine joint-probability distributions. Regarding the first criterion, we show that the Kirkwood–Dirac distributions on two- and three-state quantum systems yield faithful quasi-classical representations of quantum states with a minimal set of observables, namely a pair of two different directions of spin, and thereby point out that in general the imaginary parts of the QJP distributions play essential roles in this respect. As for the second criterion, we prove that the Kirkwood–Dirac distributions on finite-state quantum systems are supported on the product set of the spectra of the quantum observables involved.
BibTeX
@inproceedings{umekawa2024advantages,
  author = {Shun Umekawa and Jaeha Lee and Naomichi Hatano},
  title = {{Advantages of the Kirkwood–Dirac distribution among general quasi-probabilities on finite-state quantum systems}},
  booktitle = {Progress of Theoretical and Experimental Physics},
  year = {2024},
  doi = {10.1093/ptep/ptae005},
}

Loading related content...

Figure