Random-matrix models of monitored quantum circuits

Vir B. Bulchandani,S. L. Sondhi,J. T. Chalker

DOI: https://doi.org/10.1007/s10955-024-03273-0

2024-05-05

Abstract:We study the competition between Haar-random unitary dynamics and measurements for unstructured systems of qubits. For projective measurements, we derive various properties of the statistical ensemble of Kraus operators analytically, including the purification time and the distribution of Born probabilities. The latter generalizes the Porter-Thomas distribution for random unitary circuits to the monitored setting and is log-normal at long times. We also consider weak measurements that interpolate between identity quantum channels and projective measurements. In this setting, we derive an exactly solvable Fokker-Planck equation for the joint distribution of singular values of Kraus operators, analogous to the Dorokhov-Mello-Pereyra-Kumar (DMPK) equation modelling disordered quantum wires. We expect that the statistical properties of Kraus operators we have established for these simple systems will serve as a model for the entangling phase of monitored quantum systems more generally.

Quantum Physics,Disordered Systems and Neural Networks,Statistical Mechanics,High Energy Physics - Theory

What problem does this paper attempt to address?

This paper aims to explore the dynamical behavior of quantum circuits when affected by random measurements. Specifically, the competitive relationship between Haar - random unitary evolution and measurement is studied, especially in unstructured qubit systems. The main focuses include:

1. **Dynamical Purification**: The process of how a quantum state gradually changes from a mixed state to a pure state under the action of projective measurements, as well as the time scale of this process - the purification time, is studied. The analytical expression of the purification time is derived in the paper, and its variation with the system size is discussed.

2. **Born Probability Distribution**: For random unitary circuits, in the case of introducing measurements, the distribution of Born probabilities is derived. This distribution is a generalization of the classical Porter - Thomas distribution and is applicable to quantum circuits in a monitoring setting. In particular, the paper proves that in the long - time limit, the Born probability distribution is log - normal.

3. **Weak Measurement Model**: In addition to projective measurements, weak measurements (which are between the identity quantum channel and projective measurements) are also considered. By introducing a Fokker - Planck equation similar to the Dorokhov - Mello - Pereyra - Kumar (DMPK) equation, the authors are able to solve exactly the joint singular - value distribution function, revealing the "semicircle - to - square" crossover phenomenon from log - GUE statistics to log - normal statistics.

### Formula Summary

- **Lyapunov Exponent**:
\[
\lambda_n=-\frac{1}{2}(\psi(N - n + 1)-\psi(M - n + 1)),\quad n = 1,2,\ldots,M
\]
where \(\psi(x)=\frac{\Gamma'(x)}{\Gamma(x)}\) is the digamma function.

- **Purification Time**:
\[
\tau_P^{-1}=2(\lambda_1-\lambda_2)=\frac{1}{M - 1}-\frac{1}{N - 1}
\]

- **Born Probability Distribution**:
\[
p(m)\sim\prod_{j = 1}^tY_j,\quad Y_j\sim\text{Beta}(M,N - M)
\]
\[
\log p(m)\sim\sum_{j = 1}^t\log Y_j
\]
In the long - time limit, the Born probability distribution is log - normal:
\[
\log p(m)\to\mu t+\sigma t^{1/2}Z_t,\quad t\to\infty
\]
where
\[
\mu=\psi(M)-\psi(N)<0,\quad\sigma^2=\psi'(M)-\psi'(N)>0
\]
\(Z_t\sim N(0,1)\) is a standard normal random variable.

- **Rényi Entropy**:
\[
S_m^{(\alpha)}(t)=\frac{1}{1 - \alpha}\log\left(\sum_{n = 1}^M\left(\frac{\sigma_n^2(t)}{\sum_{n'= 1}^M\sigma_{n'}^2(t)}\right)^\alpha\right)
\]
In the short - time limit, the Rényi entropy is mainly determined by the moments of the average spectral density:
\[
\bar{S}^{(\alpha)}(t)=\frac{1}{1 - \alpha}\log\left(\frac{m(\alpha)}{(m(1))^{\alpha}}\right)