Permutation-symmetric quantum trajectories

This paper introduces a permutation-symmetric stochastic unraveling method that drastically reduces the computational cost of simulating exact quantum dynamics for $N$ emitters coupled to a common system, enabling efficient modeling of large system sizes for both 2-level and $d$-level emitters.

The Gist

Imagine you are trying to predict the weather for a city with a million people. If you tried to track the exact mood, location, and actions of every single person individually, the math would be so massive that even the world's fastest supercomputers would crash. This is the problem physicists face when simulating "quantum systems" made of many identical parts (like atoms in a laser or a cavity).

This paper presents a clever new way to do the math that avoids tracking every single person individually, while still getting the right answer. Here is the breakdown using everyday analogies.

The Problem: The "Crowd" vs. The "Individual"

In quantum physics, we often study groups of identical atoms (emitters) interacting with a shared environment (like a light beam in a cavity).

• The Old Way (Density Matrix): To simulate this, scientists used to treat the system like a giant spreadsheet where every possible combination of atom states was listed. If you added just one more atom, the spreadsheet grew exponentially. It was like trying to count every possible handshake in a stadium; it becomes impossible very quickly.
• The "Weak Symmetry" Problem: If all atoms were perfectly identical and behaved exactly the same, scientists could use a shortcut (treating them as one big "super-atom"). However, in the real world, atoms also lose energy individually (like a person getting tired). This "individual dissipation" breaks the perfect symmetry, forcing scientists back to the slow, impossible spreadsheet method.

The Solution: The "Stochastic Unraveling" Trick

The authors found a way to simulate these systems by looking at one possible story (a "trajectory") at a time, rather than the average of all stories.

Think of it like this:

• The Old Way: Trying to calculate the average path of a million raindrops falling through a storm.
• The New Way: Simulating the path of a single raindrop. But here is the catch: usually, if you simulate one drop, you lose the information about the crowd.

The authors discovered a special "rulebook" for how to simulate that single raindrop so that it still respects the rules of the crowd. They realized that even though individual atoms lose energy, the group's overall "shape" (mathematically called the total spin) stays predictable between random "jumps" of energy loss.

The Analogy: The Chorus Line

Imagine a choir of 1,000 singers.

1. Perfect Symmetry: If they all sing the exact same note at the exact same time, you only need to track one "super-singer."
2. The Problem: If individual singers start coughing or getting tired (individual dissipation), they fall out of sync. Usually, you'd have to track every single cougher to know what the choir sounds like.
3. The Paper's Trick: The authors realized that even if singers cough, the overall volume and pitch of the choir only change in specific, predictable ways. They developed a method to simulate the choir by tracking just the "group volume" and the "group pitch," only stopping to check individual coughs when absolutely necessary.

The Results: From a Mountain to a Hill

By using this method, the authors achieved massive speedups:

• For 2-level systems (simple atoms): They reduced the computer work from something that grows like a mountain ($N^5$) to something that grows like a gentle hill ($N$).
  • Analogy: Instead of needing a library of books to solve a problem for 1,000 atoms, they now only need a single notebook.
• For 3-level systems (more complex atoms): They showed this works for more complex atoms too, allowing simulations of systems that were previously impossible to calculate exactly.

Why This Matters (According to the Paper)

The paper claims this allows scientists to run exact simulations on much larger systems than ever before.

• Before: They could only simulate small groups of atoms exactly. For larger groups, they had to use "approximations" (guessing the average behavior), which might miss important details.
• Now: They can simulate thousands of atoms exactly. This lets them check if those old "approximations" were actually correct.

Summary

The paper is a new mathematical "shortcut" for simulating groups of identical quantum particles. It allows scientists to simulate the behavior of huge crowds of atoms by tracking the "group mood" rather than every individual's "mood," making it possible to solve problems that were previously too big for any computer to handle.

Technical Summary

Technical Summary: Permutation-Symmetric Quantum Trajectories

Problem Statement
Simulating the exact quantum dynamics of $N$ identical emitters coupled to a common system (e.g., a cavity mode) is computationally prohibitive for large $N$. While models with strong permutation symmetry (where all terms are sums of single-emitter operators) can be reduced to collective spin operators with a cost scaling as $O(N^2 N_c^2)$ (where $N_c$ is the cavity truncation), many physically relevant models include individual dissipation terms (e.g., local dephasing or decay). These terms break the strong symmetry, preventing the use of collective spins. Standard approaches for such "weakly" symmetric systems using permutation-symmetric density matrices scale as $O(N^3 N_c^2)$. Furthermore, direct stochastic unraveling (quantum trajectory) methods typically fail to exploit this symmetry; naive unraveling of individual dissipation breaks permutation symmetry, causing the computational cost to revert to exponential scaling in $N$.

Methodology
The authors propose a stochastic unraveling of the master equation that respects weak permutation symmetry by formulating the dynamics in terms of collective spin representations.

1. Pseudo-State Formulation: Instead of tracking the full density matrix $\rho$, the method tracks a "pseudo-state" $\tilde{\rho}$ defined within the space of permutation-symmetric density matrices. For two-level systems (2LS), this space is spanned by collective spin states $|J, M\rangle\langle J, M'|$, where $J$ is the total spin and $M$ is the z-magnetization.
2. Effective Jump Operators: The authors derive effective collective jump operators, $\hat{L}_{\hat{X}, \tau}$, which describe the action of individual dissipation terms on the pseudo-state. These operators map the state $|J, M\rangle$ to $|J+\tau, M+\sigma\rangle$ with specific coefficients $f_{\hat{X}}(N, J, M, \tau, \sigma)$ derived from the properties of collective spin operators. Crucially, these operators ensure the density matrix remains block-diagonal in $J$.
3. Hybrid Unraveling: The simulation evolves a pure state $|\tilde{\psi}\rangle$ using a hybrid approach:
   • Individual Dissipation: Modeled via quantum jumps using the effective operators $\hat{L}_{\hat{\ell}, \tau}$.
   • Collective Dissipation (Cavity Loss): Modeled via quantum jumps (or diffusion, see below) using operators $\hat{C}_\alpha$.
   • Non-Jump Evolution: Governed by an effective Hamiltonian $\hat{H}_{\text{eff}}$ which is permutation symmetric and block-diagonal in $J$.
4. Refinements for Scaling:
   • U(1) Symmetry: For models with weak U(1) phase symmetry (e.g., Tavis-Cummings), the total excitation number is conserved between jumps, allowing the elimination of the cavity state index $n_c$ between jumps, reducing scaling to $O(N)$.
   • Displaced Field: For models without U(1) symmetry (e.g., Dicke model), the authors employ a hybrid unraveling using quantum state diffusion (QSD) for cavity loss combined with quantum jumps for individual dissipation. By applying a unitary displacement transform ($\hat{a} \to \hat{a} + \alpha$) to the cavity field, the expectation value $\langle \hat{a} \rangle$ is non-zero, allowing the cavity truncation $N_c$ to be kept small and constant.
5. Extension to d-Level Systems: The method is generalized to $d$-level systems using Young diagrams ($\nu$) and standard Weyl tableaux ($W_\nu$) to label the permutation-symmetric states. The computational cost scales as $O(N^{d(d-1)/2})$ (with $N_c$ eliminated via the displacement method), compared to $O(N^{d^2-1} N_c^2)$ for density matrix approaches.

Key Results

• Computational Scaling: For 2LS systems with individual dissipation, the proposed method reduces the computational cost from $O(N^5)$ (standard density matrix) to $O(N)$ (with refinements). The basic quantum jump approach scales as $O(N^2)$ due to the time-step constraint $dt \propto 1/N$ at large $N$, but the refined methods achieve linear scaling.
• Dicke Model Simulations: Simulations of the Dicke model with individual dephasing and loss were performed for system sizes up to two orders of magnitude larger than previously possible ($N \sim 10^4$). The results confirm that exact numerics approach mean-field/cumulant predictions at large $N$, but the convergence is slow.
• Tavis-Cummings Model: Using the U(1) symmetry refinement, simulations reached $N=10^4$, confirming the slow approach of exact results to cumulant predictions.
• Three-Level Systems: The method was applied to a three-level lasing model (inversionless lasing), enabling simulations up to $N \sim 100$ (depending on $d$), a regime inaccessible to standard density matrix methods which are limited to $d \le 3$ for much smaller $N$. The results show convergence to mean-field behavior.

Significance and Claims
The paper claims that this approach significantly expands the range of system sizes for which exact quantum dynamics of $N$-emitter systems with individual dissipation can be modeled. By respecting weak permutation symmetry within a stochastic unraveling, the authors demonstrate that the computational barrier of exponential scaling can be overcome, reducing it to polynomial scaling ($O(N)$ for 2LS, $O(N^{d(d-1)/2})$ for $d$-level systems).

The authors emphasize that this speedup relies on a specific prescription for unravelling (mapping individual dissipation to effective collective operators and using quantum jumps). They note this as an example of how different forms of unravelling can lead to different computational complexities, even for the same ensemble-averaged physics. The method allows for rigorous testing of approximate numerical methods (such as mean-field theory, cumulant expansions, and BBGKY hierarchies) in regimes where $N$ is large enough for mean-field approximations to be expected to hold, yet small enough that exact quantum correlations might still be relevant. The paper suggests future extensions to lattice models with weak translational symmetry and non-Markovian stochastic unravelings.