Higher-order Symmetric Quantum Mpemba Effect in Fragmented Systems

This paper demonstrates that the quantum Mpemba effect persists in strongly fragmented systems with charge and dipole conservation, manifesting as a higher-order phenomenon where asymmetries relax on distinct timescales through a mechanism involving frozen memory in inactive Krylov sectors and active relaxation in dynamic fragments.

The Gist

The Big Idea: The "Quantum Mpemba Effect"

You might have heard of the Mpemba effect in the real world: the counter-intuitive idea that hot water can sometimes freeze faster than cold water.

In the quantum world, scientists discovered a similar phenomenon called the Quantum Mpemba Effect. Imagine you have two quantum systems (like a group of tiny spinning magnets). One is "very broken" (highly disordered), and the other is "slightly broken" (closer to order). Usually, you'd expect the slightly broken one to fix itself faster. But in this effect, the very broken one actually fixes itself faster, crossing paths with the other one on its way to becoming perfect.

The New Twist: A City with Locked Neighborhoods

The authors of this paper asked a big question: Does this "hot water freezes faster" trick still work if the quantum system is stuck in a very specific, rigid structure?

They studied systems where the rules of physics are so strict that the "universe" of possible states gets chopped up into millions of tiny, disconnected islands. They call this Hilbert-space fragmentation.

The Analogy:
Imagine a giant city where the roads are blocked off.

• Frozen Neighborhoods: Some parts of the city are completely locked down. If you live there, you can't move at all. You are stuck exactly where you started.
• Active Neighborhoods: Other parts of the city have open roads. People can move around, mix, and eventually reach a peaceful, organized state.

The paper asks: If you start with a chaotic crowd in this city, will the "very chaotic" group still organize faster than the "slightly chaotic" group, even though half the city is locked down?

What They Found: A "Higher-Order" Effect

The answer is yes, but with a twist. They found a "Higher-Order Symmetric Quantum Mpemba Effect."

In these rigid systems, there are two different rules the system tries to follow:

1. Charge Conservation: Keeping the total number of "up" and "down" spins balanced.
2. Dipole Conservation: Keeping the positions of those spins balanced (not just the count, but where they are).

The researchers discovered that the system fixes these two problems on different schedules:

• The "Charge" problem gets fixed first (or crosses over first).
• The "Dipole" problem gets fixed later.

It's like a runner who fixes their shoes (Charge) quickly, but then has to stop and tie their shoelaces (Dipole) much later. Both happen, but at different times.

How It Works: The "Frozen Memory" vs. The "Active Fix"

The paper explains why this happens by looking at the two types of neighborhoods mentioned earlier:

1. The Frozen Fragments (The Memory):
   In the locked-down neighborhoods, the particles are stuck. They remember exactly how "broken" they were at the start. They never fix themselves. This creates a "floor" of imperfection that never goes away. It's like a group of people who are stuck in a room and can't leave; they stay messy forever.

2. The Active Fragments (The Fixers):
   In the open neighborhoods, the particles can move. Here is where the magic happens. The group that started out more chaotic actually moves faster to fix itself than the group that started out less chaotic. They cross paths and become more ordered than the other group for a while.

The Result: The system is a mix of these two. The "Active" part rushes to fix things (creating the Mpemba crossing), while the "Frozen" part stays messy forever (creating a permanent plateau of imperfection).

How They Proved It

The authors didn't just guess; they used three different ways to test this:

1. Random Circuits: They simulated a giant, random quantum computer (up to 128 spins) using a special math trick called a "Replica Tensor Network" to see how the chaos evolved.
2. Hamiltonian Dynamics: They used a specific, non-random set of physics rules (a "pair-hopping" machine) to show this isn't just a fluke of randomness.
3. A Simple Toy Model: They built a tiny, solvable model with a "bath" (like a noisy environment) that acts as a dephaser. This allowed them to write down a perfect mathematical formula proving that the "crossing" happens and calculating exactly when it happens.

The Bottom Line

This paper shows that even in the most rigid, broken-up quantum systems, the "Mpemba effect" (where the worse-off state recovers faster) still exists. However, the rigid structure splits the recovery into two parts:

• Active parts that race to fix the symmetry (causing the crossing).
• Frozen parts that keep a permanent memory of the initial mess.

It turns out that being "more broken" can actually give you a head start in fixing the parts of the system that are allowed to move, even if the rest of the system stays stuck forever.

Technical Summary

Technical Summary: Higher-Order Symmetric Quantum Mpemba Effect in Fragmented Systems

Problem Statement
The Quantum Mpemba Effect (QME) describes an anomaly where a quantum system initially further from equilibrium (possessing more of a specific resource, such as symmetry breaking) relaxes faster than a system initially closer to equilibrium. While the QME has been established in various contexts, including integrable models and random circuits with standard $U(1)$ charge conservation, its behavior in systems with Hilbert-space fragmentation (HSF) remains an open question. HSF occurs in constrained systems (e.g., those conserving both charge and dipole moments) where the Hilbert space decomposes into exponentially many dynamically disconnected Krylov sectors. These sectors include "frozen" states that do not evolve and "active" sectors that relax slowly. The central problem addressed is whether the QME survives in such fragmented regimes and how the coexistence of frozen and active dynamics reshapes the symmetry restoration process.

Methodology
The authors investigate this problem using three complementary settings, all featuring simultaneous charge ($Q$) and dipole ($P$) conservation:

1. Charge- and Dipole-Conserving Random Circuits:

   • Model: A spin-1/2 chain evolving under a brickwork circuit where gates are drawn from the unitary group preserving both $Q$ and $P$. The minimal nontrivial gate spans four sites, allowing only specific pair-hopping moves.
   • Method: To overcome the limitations of exact vector simulation (restricted to small $L$), the authors employ a Replica Tensor Network (RTN) formulation. They adapt the RTN to charge- and dipole-conserving gates to compute the ensemble-averaged (annealed) Rényi-2 entanglement asymmetry. This allows simulations up to system sizes of $L=128$.
   • Analysis: The dynamics are resolved by projecting the state onto frozen and active Krylov sectors to isolate their respective contributions.

2. Charge- and Dipole-Conserving Pair-Hopping Hamiltonian:

   • Model: A coherent, continuous-time evolution governed by a Hamiltonian constructed from the same elementary pair-hopping processes as the circuit.
   • Method: Exact vector simulation using Chebyshev polynomial expansion for $L \le 20$. The von Neumann entanglement asymmetry is computed directly.
   • Analysis: Comparison of boundary versus bulk subsystems to test the robustness of the effect against boundary artifacts.

3. Symmetric Pair-Flipping Dissipative Toy Model:

   • Model: A minimal model with symmetric pair-flip interactions coupled to local on-site dephasing (Lindblad dynamics). This model factorizes into independent reflection-symmetric pairs.
   • Method: Exact analytical solution. The dynamics factorize, allowing closed-form expressions for the charge and dipole asymmetries.
   • Role: This model isolates the mechanism of relaxation and provides an exact benchmark for the crossing times.

Key Contributions and Results

• Discovery of Higher-Order QME: The paper establishes that the QME persists in fragmented systems but acquires a higher-order structure. Both charge and dipole asymmetries exhibit Mpemba-like crossings (where the initially more asymmetric state relaxes faster), but they do so on parametrically distinct timescales.

  • In the random circuit ($L=128$), crossing times scale as $t_M^Q \sim L_A^{2.09}$ for charge and $t_M^P \sim L_A^{1.49}$ for dipole for small subsystems.
  • In the dissipative toy model, the first crossing time is analytically derived as $t_M = \arcsin[1/\sqrt{\sin^2 \theta_1 + \sin^2 \theta_2}]$, independent of subsystem size and the specific symmetry ($Q$ or $P$) at leading order.

• Mechanism: Frozen Memory vs. Active Relaxation:

  • The authors resolve the dynamics into frozen and active Krylov sectors.
  • Frozen Fragments: These retain a finite, non-decaying asymmetry that acts as a "memory" of the initial symmetry breaking. They obstruct full symmetry restoration, leading to a late-time plateau in the asymmetry.
  • Active Fragments: These host the relaxation dynamics. The QME crossings occur entirely within the active sector, where the initially more asymmetric state relaxes faster.
  • Conclusion: Fragmentation does not preclude the QME; rather, it reshapes it into a competition between a frozen channel (preventing full restoration) and an active channel (driving the anomalous crossings).

• Robustness Across Models:

  • The effect is observed in both stochastic (random circuits) and deterministic (Hamiltonian) evolutions.
  • In the Hamiltonian setting, the effect persists for both boundary and bulk subsystems, though bulk subsystems show weaker plateaus, suggesting boundary-localized frozen structures play a significant role in the magnitude of the obstruction.
  • The dissipative toy model confirms that dephasing is necessary for monotonic relaxation toward symmetry restoration; without it, the system exhibits persistent oscillations without a clear Mpemba crossing signature.

Significance
The paper provides a systematic framework for understanding the Quantum Mpemba Effect in the presence of higher-moment symmetries and strong Hilbert-space fragmentation. It demonstrates that the QME is a robust phenomenon that adapts to the constraints of the system, splitting symmetry restoration into distinct frozen and active components. This "frozen-plus-active" decomposition offers a new perspective on how constrained quantum systems store and erase information: frozen fragments protect long-lived symmetry-breaking memory, while active fragments facilitate relaxation. The findings suggest that the higher-order QME can serve as a diagnostic tool for probing memory lifetimes and relaxation bottlenecks in constrained quantum simulators, such as Rydberg arrays and trapped ions, where dipole conservation can be engineered.