Computing Large Deviations of First-Passage-Time Statistics in Open Quantum Systems: Two Methods

This paper proposes and validates two distinct methods—one analytical based on pole analysis of transformed distributions and one simulation-based using wave function cloning—for computing the large deviations of first-passage-time statistics in general open quantum systems.

The Gist

Imagine you are watching a busy train station. You want to know two things about the passengers:

1. Counting Statistics: "How many people have passed through the turnstile after 1 hour?"
2. First-Passage-Time (FPT) Statistics: "How long does it take for exactly 100 people to pass through the turnstile?"

In the world of Open Quantum Systems (tiny particles interacting with their environment, like atoms in a laser beam), these questions are incredibly hard to answer. The particles jump around randomly, and the math to predict their behavior is usually a nightmare of complex equations.

This paper by Fei Liu and Jiayin Gu proposes two new, clever ways to solve these "rare event" puzzles without getting lost in the math.

Here is the breakdown using simple analogies:

The Big Problem: The "Impossible" Math

Usually, to predict these rare events, scientists have to solve a giant equation that describes the "tilted" behavior of the system.

• The Old Way: It's like trying to find the exit of a massive, shifting maze by looking at a map of the entire maze at once. For simple systems, it works. But for complex systems (like a crowd of interacting atoms), the map is so huge it crashes your computer.
• The Insight: The authors realized that the math for "Counting" (how many jumps) and "First-Passage-Time" (how long it takes) are actually mirror images of each other. If you know one, you can mathematically flip it to get the other. But they wanted to find ways to solve the "Time" problem directly, just in case the "Counting" problem is too hard to solve first.

Method 1: The "Pole Hunter" (The Map Reader)

This method is for systems that aren't too huge.

• The Analogy: Imagine the behavior of the quantum system as a radio signal. Most signals are clear, but there are specific "dead zones" or "poles" where the signal breaks down. These dead zones actually hold the secret to the answer.
• How it works: Instead of trying to simulate every single jump, the authors found a specific equation (a "pole equation") that tells you exactly where these dead zones are.
  • If you solve this equation, you find the boundary of a safe zone.
  • The edge of this safe zone gives you the answer directly.
• The Catch: This is like trying to find the edge of a maze by looking at a 2D map. It works great for a small maze (a single atom), but if the maze is a 3D city (many interacting atoms), the map becomes too big to draw, and this method gets stuck.

Method 2: The "Wave Function Cloning" (The Army of Simulations)

This method is for the massive, complex systems where Method 1 fails.

• The Analogy: Imagine you want to know how long it takes for a specific rare event to happen in a crowd. Instead of watching one person for a million years, you create 10,000 clones of that person and watch them all at once.
  • The Twist: If a clone is doing something "unlikely" (like taking a very long time to reach the goal), you kill that clone.
  • If a clone is doing something "likely" (moving fast), you clone it (make a copy of it).
  • You keep the total number of clones constant by randomly adding or removing them.
• How it works: By constantly pruning the slow paths and multiplying the fast paths, the simulation naturally focuses on the "rare" events you care about. It's like a game of "survival of the fittest" for computer simulations.
• The Result: Even though the system is huge, this "cloning army" efficiently finds the answer by ignoring the boring, average paths and zooming in on the rare, extreme ones.

The "Test Drive" (Examples)

To prove their methods work, the authors tested them on three scenarios:

1. A Single Atom (Two-Level System): Like a light switch that is either ON or OFF. They solved it with Method 1 and got a perfect match with known results.
2. A Three-Level Atom: A slightly more complex switch. Again, Method 1 worked beautifully, giving them a neat formula.
3. Two Atoms Talking to Each Other: This is the hard one. The atoms interact, making the "map" too complex for Method 1. Here, they used Method 2 (Cloning). The simulation successfully predicted the statistics, proving that even for complex, interacting systems, you can find these rare time statistics.

Why Does This Matter?

In the quantum world, "rare events" are often the most important ones. They determine things like:

• How much energy a quantum computer might leak.
• How long a quantum sensor stays accurate before noise ruins it.
• The fundamental limits of how fast information can be processed.

In summary: This paper gives scientists two new tools. One is a mathematical shortcut (finding the poles) for simple systems, and the other is a smart simulation trick (cloning) for complex systems. Together, they allow us to predict the timing of rare quantum events with much greater ease and accuracy than before.

Technical Summary

1. Problem Statement

The paper addresses the challenge of computing Large Deviation Functions (LDFs), specifically the Scaled Cumulant Generating Functions (SCGFs), for First-Passage-Time (FPT) statistics in general open quantum systems.

• Context: While large deviations in counting statistics (time-extensive variables) are well-studied, FPT statistics (the time required for a counting variable to reach a specific threshold) are distinct and mathematically more complex.
• Existing Limitations: Previous methods relied heavily on the "inverse function relationship" between the SCGFs of counting statistics and FPT statistics. However, this relationship is not always easy to utilize, and computing the SCGF for counting statistics itself is often computationally intractable for complex systems. Furthermore, many open quantum systems cannot be described by semi-Markov processes (a common simplification), rendering previous analytical methods inapplicable.
• Goal: To develop two independent, general methods for calculating FPT large deviations in open quantum systems without relying solely on the inverse relationship with counting statistics.

2. Methodology

The authors propose two distinct approaches based on the piecewise deterministic process (PDP) of wave functions and the tilted Liouville master equation.

Method 1: Analytical Pole Equation Approach

This method extends previous work on semi-Markov processes to general open quantum systems.

• Theoretical Basis: The dynamics of the system are unraveled into quantum jump trajectories. The authors utilize the tilted Liouville master equation in Hilbert space to describe the joint probability distribution of the wave function and the counting variable.
• Key Derivation:
  • They establish a relationship between the joint transition probability and the FPT probability distribution in the frequency domain (Laplace transform).
  • They identify that the boundary of the Region of Convergence (ROC) for the joint Laplace and $z$-transforms is determined by the poles of the tilted generator.
  • They derive an algebraic equation of poles: $\det[\nu - \tilde{L}_z] = 0$, where $\nu$ is the Laplace variable (conjugate to time) and $z$ is the $z$-transform variable (conjugate to the counting variable).
• Procedure:
  1. Construct the tilted generator $\tilde{L}_z$ in a specific representation (e.g., energy basis).
  2. Solve the characteristic polynomial equation $\det[\nu - \tilde{L}_z] = 0$ for roots.
  3. Identify the boundary curve in the real $\nu$-$z$ plane.
  4. Extract the SCGFs:
     • For counting statistics: $\Phi(\lambda) = \max_k \text{Re}[\nu_k(z)]$ where $z = e^\lambda$.
     • For FPT statistics: $\Psi(\nu) = \ln(\max_m |z_m(\nu)|)$, where $z_m(\nu)$ are the roots of the pole equation.

Method 2: Wave Function Cloning Algorithm (Simulation)

This is a numerical approach designed for systems with large Hilbert spaces where analytical solutions are impossible.

• Concept: A variant of the cloning algorithm (originally used for counting statistics) adapted for FPT statistics.
• Mechanism:
  • Simulates a population of quantum systems evolving via quantum jump trajectories.
  • Cloning/Pruning: Based on a cloning factor derived from the conjugate parameter $\nu$ (for time) and the jump weights, systems are either terminated or cloned to bias the ensemble toward rare events.
  • FPT Adaptation: Unlike counting statistics (where time is fixed and the variable grows), FPT statistics involves a stochastic variable (the count) reaching a fixed threshold. The algorithm loops until all system copies reach a predetermined threshold $n$.
  • Handling Current-like Variables: For variables that can increase or decrease (current-like), the algorithm handles two branches ($\Psi_+$ and $\Psi_-$) corresponding to positive and negative thresholds.

3. Key Contributions

1. Generalization of Pole Equations: The authors successfully extend the "equation of poles" method from semi-Markov processes to general open quantum systems described by the tilted Liouville master equation in Hilbert space. This removes the restriction that systems must be describable as semi-Markov processes.
2. Unified Treatment: The analytical method provides a unified framework for both simple counting variables (non-negative, e.g., dynamic activity) and current-like variables (can be positive or negative, e.g., entropy production).
3. Independence from Inverse Relationship: The methods are derived directly from FPT statistics principles, proving that one can compute FPT large deviations without first solving the more difficult counting statistics problem and applying an inverse transform.
4. Algorithmic Innovation: The development of a specific wave function cloning algorithm for FPT statistics that correctly handles the stochastic nature of the counting variable as the "time" parameter in the simulation loop.

4. Results

The authors validated both methods using three specific models:

• Field-Driven Two-Level System:

  • Derived analytical expressions for SCGFs for both current-like (entropy production) and simple (dynamic activity) variables.
  • Confirmed the inverse function relationship ($\Psi = \Phi^{-1}$) between counting and FPT statistics.
  • Verified the Fluctuation Theorem ($\Psi_+(\nu) = \beta\omega_0 - \Psi_-(\nu)$).
  • Showed that solving the pole equation is analytically simpler than solving for counting statistics in non-resonant cases.

• Field-Driven Three-Level System (V-type):

  • Derived the SCGF for dynamic activity.
  • Identified a smoothed first-order dynamical phase transition at $\nu=0$ when parameters approach zero.
  • Demonstrated that the FPT SCGF terminates at the origin in the limit of a true dynamical transition, indicating a breakdown of the large deviation principle (infinite tail in the distribution).

• Two Interacting Two-Level Atoms:

  • This system cannot be described by a semi-Markov process due to memory effects in the collapsed wave function.
  • The pole equation resulted in a high-degree polynomial (16th degree in $\nu$, 10th in $z$), making analytical solution difficult.
  • The cloning simulation successfully computed the SCGFs.
  • Results from the simulation perfectly matched the numerical roots of the pole equation, validating the simulation method for complex, interacting many-body systems.

5. Significance

• Theoretical Completeness: The paper establishes that FPT statistics in open quantum systems has its own rigorous theoretical foundation, independent of counting statistics, though they are mathematically linked.
• Computational Efficiency: For systems with large Hilbert spaces (many-body systems), the cloning algorithm is presented as an essential tool, as solving high-dimensional pole equations analytically is often impossible.
• Physical Insight: The ability to directly compute FPT large deviations allows for the study of rare events, uncertainty relations, and dynamical phase transitions in quantum systems that were previously inaccessible or required indirect, approximate methods.
• Broad Applicability: The methods apply to any open quantum system governed by a Lindblad master equation, covering both simple and current-like observables, making them highly relevant for quantum thermodynamics, quantum transport, and quantum metrology.