Quantum Mpemba Symmetry Restoration

---
name: quantum-mpemba-symmetry-restoration
description: "Quantum Mpemba effect methodology for symmetry restoration in fragmented Hilbert spaces. Covers higher-order symmetric quantum Mpemba effect where quantum systems restore broken symmetry faster the more strongly it's initially broken. Uses replica tensor-network formulation for charge/dipole-conserving gates, frozen vs active Krylov sector decomposition, and Rényi-2 entanglement asymmetry analysis. Applicable to quantum dynamics, non-equilibrium quantum systems, memory retention in fragmented systems. Activation: quantum mpemba effect, symmetry restoration, hilbert space fragmentation, krylov sectors, entanglement asymmetry, replica tensor network, quantum memory dynamics, non-equilibrium quantum."
category: quantum-physics
---

# Quantum Mpemba Symmetry Restoration in Fragmented Systems

Methodology from "Higher-order Symmetric Quantum Mpemba Effect in Fragmented Systems" (arXiv:2606.06653). Sreemayee Aditya, Sara Murciano, Xhek Turkeshi.

## Core Insight

A quantum system can restore a broken symmetry faster the more strongly it initially breaks it — the quantum Mpemba effect. This effect persists even when conservation laws fragment the Hilbert space into exponentially many disconnected Krylov sectors (charge + dipole conservation).

## Key Discovery: Higher-Order Symmetric Quantum Mpemba Effect

### Fragmentation Does Not Destroy Mpemba — It Reshapes It

The Mpemba effect survives Hilbert space fragmentation but transforms into two distinct mechanisms:

1. **Frozen fragments**: Retain a finite asymmetry that obstructs full symmetry restoration — analogous to "frozen memory" in neural systems
2. **Active fragments**: Host the relaxation responsible for Mpemba crossings — the mechanism that drives the anomalous faster restoration

### Dual-Timescale Mpemba Crossings

- **Charge asymmetry** displays Mpemba-like crossings on one parametric timescale
- **Dipole asymmetry** displays Mpemba-like crossings on a parametrically distinct timescale
- This reveals the Mpemba phenomenology of higher-moment symmetries

## Core Methodology

### 1. Replica Tensor-Network Formulation

- Develop replica tensor networks for charge and dipole-conserving gates
- Reach annealed Rényi-2 entanglement asymmetry up to L=128 system sizes
- Compute symmetry-resolved entanglement dynamics in fragmented Hilbert spaces

### 2. Krylov Sector Decomposition

- Resolve the quantum state into frozen and active Krylov sectors
- Frozen sectors: exponentially many disconnected subspaces that retain finite asymmetry
- Active sectors: subspaces where relaxation dynamics occur and Mpemba crossings emerge
- **Key insight**: Fragmentation reshapes Mpemba into "frozen memory" + "active-fragment relaxation"

### 3. Complementary Simulation Approaches

- **Circuit simulations**: Using replica tensor-network formulation for gate-based dynamics
- **Hamiltonian simulations**: Direct time evolution under fragmented Hamiltonians
- **Exactly solvable dissipative model**: Analytical solution for Mpemba effect validation

### 4. Quantification via Entanglement Asymmetry

- Use Rényi-2 entanglement asymmetry as the order parameter for symmetry restoration
- Track asymmetry evolution across different initial symmetry-breaking strengths
- Identify Mpemba crossings (where stronger initial breaking leads to faster restoration)

## Implementation Steps

1. **Define fragmented system**: Set up charge + dipole conserving circuit/Hamiltonian
2. **Construct replica tensor network**: Implement charge/dipole-conserving gates in tensor format
3. **Decompose into Krylov sectors**: Identify frozen vs active subspaces
4. **Compute Rényi-2 asymmetry**: Track symmetry restoration dynamics per sector
5. **Identify Mpemba crossings**: Compare restoration times for different initial conditions
6. **Validate with dissipative model**: Cross-check against exactly solvable limit

## Pitfalls

- **Exponential fragmentation**: Hilbert space fragmentation creates exponentially many sectors — direct simulation becomes intractable beyond L~20 without tensor network methods
- **Annealed vs quenched asymmetry**: Use annealed Rényi-2 for tensor network tractability; quenched averages may differ
- **Timescale separation**: Charge and dipole asymmetries relax on parametrically distinct timescales — must track both independently
- **Frozen sector trapping**: Frozen fragments can prevent full symmetry restoration even at infinite time — Mpemba effect manifests as partial restoration speedup, not complete restoration

## Verification

- Mpemba crossing should appear in both circuit and Hamiltonian simulations
- Dissipative model should reproduce the crossing behavior analytically
- Frozen sector asymmetry should remain finite at long times
- Active sector contribution should dominate the crossing dynamics

## Connection to Neuroscience

The "frozen memory" mechanism in fragmented Krylov sectors provides a quantum analogue for neural memory retention — certain patterns (frozen sectors) resist decay while others (active sectors) undergo dynamic restoration. This connects to:
- Persistent activity in working memory neural circuits
- Attractor dynamics in Hopfield networks
- Memory consolidation vs forgetting in biological systems