⚛️ quantum physics

Fermionic magic resources in disordered quantum spin chains

This paper demonstrates that fermionic non-Gaussianity, quantified by fermionic antiflatness, is suppressed in the many-body localized regime of disordered spin chains—exhibiting area-law bounds and slow power-law growth—while being restored in ergodic phases with volume-law scaling, thereby establishing it as a sensitive diagnostic for distinguishing between localization and thermalization.

Original authors: Pedro R. Nicácio Falcão, Jakub Zakrzewski, Piotr Sierant

Published 2026-02-03

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Pedro R. Nicácio Falcão, Jakub Zakrzewski, Piotr Sierant

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Picture: Magic in a Messy Kitchen

Imagine you have a kitchen full of ingredients (quantum particles). Some kitchens are very organized, where you can predict exactly what happens if you mix two things. In the quantum world, these "organized" states are called free-fermionic states. They are easy for classical computers to simulate, like following a simple recipe.

However, real quantum systems often have "interactions" (ingredients that react strangely with each other). When these interactions happen, the state becomes non-Gaussian. The authors of this paper call this "Fermionic Magic." Think of "Magic" as the amount of "quantum weirdness" or complexity that makes a system hard for a normal computer to predict.

The paper asks: What happens to this "Magic" when the kitchen is messy (disordered) and the ingredients get stuck?

The Two Scenarios: The Flowing River vs. The Frozen Pond

The researchers studied two types of quantum "kitchens" (spin chains) to see how "Magic" behaves under different conditions:

The XXZ Chain (The Whole Kitchen): Imagine a long row of pots where every pot has a slightly different amount of disorder (randomness).
The Impurity Model (One Bad Apple): Imagine the same row of pots, but only one specific spot has a strong interaction, while the rest are free.

They looked at two main regimes:

Ergodic (The Flowing River): When disorder is low, the system is fluid. Information spreads everywhere quickly.
Many-Body Localized (MBL) (The Frozen Pond): When disorder is high, the system gets "stuck." Information cannot spread; it stays trapped in small pockets.

Key Findings

1. Magic is Suppressed in the "Frozen" State

When the system enters the MBL regime (the frozen pond), the "Magic" (complexity) drops significantly.

The Analogy: Imagine trying to make a complex, swirling pattern in a frozen lake. No matter how hard you try, the ice keeps the water still. The "Magic" is suppressed because the disorder locks the particles in place, preventing them from interacting in complex ways.
The Result: In the "frozen" state, the system behaves almost like the simple, easy-to-calculate "free" states. The more disorder you add, the less "Magic" you have.

2. The Size of the "Bad Apple" Matters

The researchers found that how much "Magic" you get depends on how much of the system is interacting.

In the XXZ Chain: The interaction happens everywhere. Even in the frozen state, the "Magic" grows with the size of the system (Volume Law). It's like having a few frozen spots in a large lake; the ice is thick, but the whole lake still has some complexity.
In the Impurity Model: Only one spot interacts. In the frozen state, the "Magic" stays small and doesn't grow with the system size (Area Law). It's like having a single frozen patch in a huge lake; the rest of the lake is irrelevant to that one spot.

3. The "Ghost" of the Cat (Rare Resonances)

Sometimes, even in a frozen system, a rare event happens where two distant parts of the system suddenly "talk" to each other. The paper calls these "Cat-like eigenstates."

The Analogy: Imagine a frozen pond where, by pure chance, a giant wave suddenly forms in two distant corners simultaneously, creating a "Schrödinger's Cat" situation (both frozen and flowing at once).
The Result: These rare events are like "magic bombs." They contain a huge amount of "Magic" (non-Gaussianity) compared to the rest of the system. The authors found that detecting this high "Magic" is a great way to spot these rare, destabilizing events that might eventually break the "frozen" state.

4. Time Travel: How Fast Does Magic Grow?

The researchers watched what happens when they start with a simple, ordered state (like a neat row of up/down spins) and let time pass.

In a Normal System (Ergodic): "Magic" grows fast and saturates quickly, like a drop of ink spreading instantly in water.
In the Frozen System (MBL): "Magic" grows incredibly slowly. It's like watching a drop of ink spread through thick honey. It takes a very long time to reach its maximum complexity, and it follows a specific, slow mathematical pattern (power-law).

Summary

This paper shows that disorder acts like a dam that stops the flow of quantum complexity ("Magic").

In a frozen (MBL) system, the "Magic" is low and grows very slowly over time.
However, rare, giant "Cat-like" events can suddenly create a massive burst of "Magic," acting as a warning sign that the frozen state might be unstable.
The amount of "Magic" depends on whether the interactions are spread out (like a whole chain) or localized (like a single impurity).

The authors conclude that studying this "Fermionic Magic" helps us understand how quantum systems resist or succumb to becoming complex and chaotic.

Technical Summary: Fermionic Magic Resources in Disordered Quantum Spin Chains

Problem Statement
This work investigates the behavior of fermionic non-Gaussianity, a key resource for quantum computational advantage, in disordered many-body quantum systems. While entanglement and non-stabilizerness ("magic") have been extensively studied as diagnostics for quantum complexity and ergodicity breaking, fermionic non-Gaussianity remains less explored in many-body contexts. The central problem is to quantify how many-body localization (MBL)—a robust mechanism of ergodicity breaking—suppresses the buildup of fermionic magic resources compared to ergodic thermalizing systems. Specifically, the authors aim to characterize the deviation from free-fermionic descriptions (Gaussian states) across the ergodic-MBL crossover in highly excited eigenstates and during real-time dynamics.

Methodology
The study focuses on two paradigmatic disordered spin-1/2 models mapped to fermionic systems via the Jordan-Wigner transformation:

Disordered XXZ Chain: A model with nearest-neighbor interactions and random onsite fields, representing a transition from an Anderson-localized free-fermion limit to an interacting MBL/ergodic regime.
Impurity Model: A variant where interactions are confined to a single bond at the center of the chain, allowing for a comparison between extensive and localized interaction effects.

The authors employ Fermionic Antiflatness (FAF), denoted as $F_k$ , as the primary measure of non-Gaussianity. FAF is defined via the covariance matrix $M$ of the state:
$F_k(|\Psi\rangle) = L - \frac{1}{2}\text{tr}[(M^T M)^k]$
where $L$ is the system size. FAF is a faithful measure (zero for Gaussian states), Gaussian invariant, and additive.

The computational approach involves:

Eigenstate Analysis: Using the POLFED algorithm to compute mid-spectrum eigenstates in the zero-magnetization sector. Results are averaged over thousands of disorder realizations.
Time Evolution: Simulating real-time dynamics from N´eel product states using exact diagonalization for system sizes $L=10$ to $L=20$ .
Phenomenological Modeling: Utilizing a phenomenological description based on quasi-local integrals of motion (LIOMs or $\ell$ -bits) and a U(1)-symmetric circuit to derive analytical scaling laws for FAF dynamics.

Key Contributions and Results

Suppression of Non-Gaussianity in MBL:
The study demonstrates that MBL strongly suppresses fermionic non-Gaussianity. In the ergodic regime (weak disorder), FAF density $\langle f_1 \rangle$ approaches unity, consistent with typical-state behavior. In contrast, deep in the MBL regime (strong disorder), FAF is significantly suppressed.
- Scaling with Disorder: In the large-disorder limit, FAF scales as $\langle f_1 \rangle \propto W^{-2}$ , where $W$ is the disorder strength. This suppression arises because the interacting system remains smoothly connected to the Anderson-localized free-fermion limit via a quasi-local unitary dressing.
- Interaction Extensivity: The scaling of FAF depends on the spatial extent of interactions. The XXZ chain (extensive interactions) exhibits volume-law scaling ( $F_1 \propto L$ ) in the MBL regime, whereas the impurity model (local interactions) exhibits an area-law bound (constant $F_1$ independent of $L$ ).
Perturbative Regime and Cat-like Eigenstates:
The authors derive a perturbative expression for FAF deep in the MBL regime:
$F_k(|\Psi_{\text{MBL}}\rangle) \propto \frac{\Delta N_{\text{int}}}{W^2}$
where $\Delta$ is the interaction strength and $N_{\text{int}}$ is the number of interacting bonds.
Crucially, the study identifies rare long-range cat-like eigenstates as a mechanism that destabilizes MBL. These states, which involve macroscopic superpositions (cat states) over resonant segments, exhibit a pronounced enhancement of FAF. For these states, FAF isolates the contribution of the resonant segment, yielding values proportional to the size of the resonance. This links non-perturbative mechanisms for MBL destabilization directly to increased fermionic magic resources.
Real-Time Dynamics:
In the MBL regime, the growth of FAF from product initial states is anomalously slow compared to ergodic systems.
- Power-Law Relaxation: The time evolution of FAF follows a power-law relaxation toward a saturation value:
  $F_1(t) = F_1^{\text{MBL}} - C t^{-\beta} + O(t^{-2\beta})$
  where $F_1^{\text{MBL}}$ is an extensive plateau determined by the initial state and disorder.
- Mechanism: This slow growth is driven by the slow algebraic decay of non-local Majorana correlators, which map to non-local spin strings. The dynamics are consistent with the phenomenological $\ell$ -bit description, where off-diagonal operators decay algebraically.

Significance and Claims
The paper establishes that ergodicity breaking via MBL acts as a constraint on the generation of fermionic magic resources, keeping the system closer to a classically simulable free-fermionic description. Conversely, ergodicity restores the complexity associated with non-Gaussianity.

The authors claim that FAF serves as a sensitive diagnostic for:

Distinguishing between ergodic and MBL regimes.
Detecting rare, non-perturbative resonant states (cat-like eigenstates) that may destabilize MBL in the thermodynamic limit.
Characterizing the slow dynamics of complexity growth in localized systems.

The work suggests that the phenomenology of fermionic magic resources extends to other non-ergodic systems admitting a localized free-fermion limit, such as quasiperiodic potentials and disorder-free localization. The results provide a reference point for future studies of fermionic non-Gaussianity in other ergodicity-breaking phenomena, including many-body scars and gauge-constrained systems.