Fabio Anza,James P. Crutchfield
Abstract:Any given density matrix can be represented as an infinite number of ensembles of pure states. This leads to the natural question of how to uniquely select one out of the many, apparently equally-suitable, possibilities. Following Jaynes' information-theoretic perspective, this can be framed as an inference problem. We propose the Maximum Geometric Quantum Entropy Principle to exploit the notions of Quantum Information Dimension and Geometric Quantum Entropy. These allow us to quantify the entropy of fully arbitrary ensembles and select the one that maximizes it. After formulating the principle mathematically, we give the analytical solution to the maximization problem in a number of cases and discuss the physical mechanism behind the emergence of such maximum entropy ensembles.
What problem does this paper attempt to address?
This paper attempts to address the problem of how to uniquely select the most suitable ensemble from the countless possible pure state ensembles given a density matrix. Specifically, the authors propose the Maximum Geometric Quantum Entropy Principle, which uses the concepts of quantum information dimension and geometric quantum entropy to quantify the entropy of any ensemble and select the ensemble that maximizes the entropy.
### Background and Motivation
1. **Background**:
- The state of a system in quantum mechanics can be represented by a pure state \( |\psi\rangle \) in a Hilbert space \( \mathcal{H} \).
- To account for the uncertainty of the actual state \( |\psi\rangle \) of the system, a density operator \( \rho \) is introduced, which acts on \( \mathcal{H} \) and satisfies the conditions of linearity, semi-positivity, self-adjointness, and normalization.
- The density operator can be decomposed as \( \rho = \sum_i \lambda_i |\lambda_i\rangle \langle \lambda_i| \), where \( \lambda_i \) are the eigenvalues and \( |\lambda_i\rangle \) are the eigenvectors.
- Given a density matrix \( \rho \), there exist infinitely many different pure state ensembles \( \{p_k, |\psi_k\rangle\} \) such that \( \sum_k p_k |\psi_k\rangle \langle \psi_k| = \rho \).
2. **Motivation**:
- Since the mapping from a density matrix to a pure state ensemble is one-to-many, it is natural to ask whether there exists a meaningful criterion to uniquely select an ensemble.
- This is a typical inference problem that can be addressed using the Maximum Entropy Principle (MEP). The maximum entropy method has achieved significant empirical success in dealing with inference problems with partial knowledge and is widely applied in science and engineering.
### Solution
1. **Maximum Geometric Quantum Entropy Principle**:
- The authors propose the Maximum Geometric Quantum Entropy Principle, which uses the concepts of quantum information dimension and geometric quantum entropy to quantify the entropy of any ensemble and select the ensemble that maximizes the entropy.
- Mathematically, this principle can be stated as: given a density matrix \( \rho \), among all ensembles with a fixed information dimension \( D \), select the ensemble \( \mu_D^{ME} \) that maximizes the geometric quantum entropy \( h_D \):
\[
\mu_D^{ME} := \arg \max_{\mu \in E_D(\rho)} h_D
\]
2. **Optimization Problem**:
- The optimization problem is subject to two constraints: the ensemble must be normalized, and the expectation value of \( (Z_j)^* Z_k \) must equal \( \rho_{jk} \).
- Using the method of Lagrange multipliers, the objective function \( \Lambda[\mu, \gamma_1, \{\gamma_{jk}\}] \) is optimized:
\[
\Lambda[\mu, \gamma_1, \{\gamma_{jk}\}] := h_D[\mu] + \gamma_1 C_1 + \sum_{j,k} \gamma_{jk} C_{\rho_{jk}}
\]
- By solving the equations where the derivative of \( \Lambda \) is zero, the desired ensemble \( \mu_D^{ME} \) is obtained.
### Physical Mechanism
- The paper discusses two mechanisms that lead to the appearance of the maximum geometric quantum entropy ensemble and identifies the occurrence of these ensembles under different physical conditions.
- For example, in the case of a finite environment (\( D = 0 \)), the support set of the ensemble consists of several points, and maximum entropy can be achieved by measuring the environment in a basis independent of the Schmidt partner basis.
- In the case of full support (