Time irreversibility and entropy production in non-Hermitian Model A field theories

Here is an explanation of the paper, translated from complex physics jargon into everyday language using analogies.

The Big Picture: Why Time Seems to Flow One Way

Imagine you drop a glass of water on the floor. It shatters. You never see the shards jump back up and reassemble into a perfect glass. This is Time Irreversibility. In the world of physics, this "one-way street" is usually a sign that a system is out of equilibrium (it's not resting comfortably; it's being pushed or driven).

Usually, if you look at a single tiny particle, it's easy to see if it's being pushed. But in complex systems made of billions of particles (like a crowd of people, a flock of birds, or a chemical soup), it's very hard to tell if the whole group is "out of equilibrium" just by looking at the big picture. Sometimes, the chaos at the microscopic level hides itself, making the system look like it's in a calm, resting state, even though it's actually burning energy and moving forward in time.

The Goal of this Paper:
The authors wanted to build a "detector" to measure exactly how much time-reversal symmetry is broken in these complex systems. They wanted to answer: How "out of equilibrium" is this system, really?

The Main Characters: The "Ghost" and the "Engine"

To understand their solution, we need to meet two concepts they use:

The Hermitian Part (The Engine): Think of this as the standard rules of physics. It's like a ball rolling down a hill. It follows the path of least resistance. This part is "reversible" in a mathematical sense; if you played the movie backward, it would still look like a valid physical process.
The Non-Hermitian Part (The Ghost): This is the new, weird ingredient. It represents forces that don't follow the standard "downhill" rules. It's like a ghost pushing the ball sideways or making it spin in a way that defies normal gravity. In the real world, this happens in active matter—systems where individual parts consume energy to move, like bacteria swimming or people in a crowd pushing each other.

The Paper's Discovery:
The authors found that you can measure the "irreversibility" (the time arrow) simply by looking at the strength of this "Ghost" (the Non-Hermitian part).

The Two Tools: The "Fingerprint" and the "Fuel Gauge"

The paper uses two main ways to measure this irreversibility. Let's use a car analogy.

1. The Fluctuation-Dissipation Theorem (FDT) Violation = The "Fingerprint"

In a calm, equilibrium system (a parked car), if you push it slightly, it wiggles and settles back down in a predictable way. The relationship between how much it wiggles (fluctuation) and how much it resists (dissipation) is a perfect, known formula.

The Analogy: Imagine a perfectly balanced seesaw. If you push one side, it goes down, and the other goes up. The relationship is symmetrical.
The Break: When the "Ghost" (Non-Hermitian force) is present, the seesaw is broken. Pushing it doesn't just make it go down; it might make it spin or slide sideways. The relationship between the push and the wiggle is broken.
The Result: The authors show that this "broken fingerprint" appears linearly. If you add a little bit of the Ghost, the fingerprint breaks a little bit. If you add a lot, it breaks a lot. It's a very sensitive detector.

2. Entropy Production Rate (EPR) = The "Fuel Gauge"

Entropy is a measure of disorder or "wasted energy." In a reversible world, you could run a process backward without losing anything. In an irreversible world, you burn fuel to keep things moving.

The Analogy: Think of the "Ghost" as a misaligned engine. Even if the car looks like it's driving straight, the engine is grinding gears and burning extra fuel.
The Result: The authors found that the amount of "fuel burned" (Entropy Production) depends on the square of the Ghost's strength.
- If the Ghost is weak, the fuel burn is very weak (tiny number squared is even tinier).
- This means the "Fuel Gauge" (Entropy) is less sensitive than the "Fingerprint" (FDT). You need a stronger Ghost to see a big change in the fuel burn.

Key Takeaway: The "Fingerprint" (FDT violation) tells you the Ghost is there immediately. The "Fuel Gauge" (Entropy) tells you how much energy is being wasted, but it takes a stronger Ghost to show a big number.

The Case Study: The "Vision Cone" Crowd

To prove their math works, the authors applied it to a specific model: The Non-Reciprocal Ising Model.

The Scenario: Imagine a crowd of people (spins) on a grid. Usually, people try to agree with their neighbors (like magnets aligning).
The Twist: In this model, the people have "Vision Cones." They can only see and react to people in front of them, not behind them.
- Analogy: Imagine a line of people passing a ball. If everyone can only see the person in front, the ball moves in a specific direction. If everyone could see everyone, the ball would just bounce around randomly.
The Result: Because of this one-way vision, the system creates "traffic jams" or boundaries (interfaces) between groups of people moving in different directions.

Where does the "Fuel" burn?
The authors calculated exactly where the entropy (fuel waste) happens in this crowd:

In the middle of the crowd (Uniform phase): If everyone is moving the same way, the "Ghost" forces cancel out, and there is almost no fuel waste. The system looks calm.
At the boundaries (Interfaces): Where two groups of people moving in different directions meet, the "Vision Cone" forces clash. This is where the friction happens.
- The Surprise: The entropy production localizes (concentrates) entirely at these boundaries (the domain walls). It's like the engine is only grinding gears at the edge of the crowd, while the middle is smooth.

Why Does This Matter?

It's a Universal Rule: They proved that for a huge class of systems, the "Ghost" (Non-Hermitian part) is the only thing that breaks time symmetry in these specific models. If you remove the Ghost, the system becomes reversible again.
It Explains Active Matter: This helps us understand things like flocks of birds, bacterial swarms, or self-driving cars. These systems are constantly burning energy. The paper gives us a mathematical way to say, "Look, the energy is being wasted right here, at the edge of the flock."
A New Toolkit: They provided a "calculator" (a framework) that scientists can use to take any complex, messy system, break it down, and calculate exactly how irreversible it is, even if the system looks calm on the surface.

Summary in One Sentence

The authors created a mathematical method to prove that in complex, active systems, the "one-way-ness" of time is caused by specific non-standard forces, and that the energy wasted by this process concentrates heavily at the boundaries where different behaviors meet, rather than being spread out evenly.

Here is a detailed technical summary of the paper "Time irreversibility and entropy production in non-Hermitian Model A field theories" by Carosi et al.

1. Problem Statement

The paper addresses the fundamental challenge of quantifying time-reversal symmetry (TRS) breaking and irreversibility in macroscopic non-equilibrium systems, specifically those described by scalar field theories with non-Hermitian dynamics.

Context: While microscopic non-equilibrium systems (e.g., active matter) explicitly break TRS, emergent macroscopic behaviors often resemble equilibrium states, making it difficult to identify "non-equilibriumness" at coarse-grained scales.
Gap: Existing methods to quantify irreversibility, such as the Entropy Production Rate (EPR) and Fluctuation-Dissipation Theorem (FDT) violations, have been primarily developed for Hermitian systems or specific active models (like Model B). There is a lack of a systematic framework to handle generic non-Hermitian terms in non-conserved (Model A) scalar field theories.
Goal: To develop a unified framework that links non-Hermitian forces to FDT violations and EPR, demonstrating how these quantities scale with the degree of non-Hermiticity.

2. Methodology

The authors employ a stochastic path-integral formalism (Martin-Siggia-Rose-Janssen-De Dominicis or MSRJD) combined with a controlled small-noise expansion.

Model Setup: They consider a real scalar field $\psi$ evolving under Model A dynamics (non-conserved order parameter) governed by a Langevin equation:
$\dot{\psi} = \Gamma F[\psi] + \sqrt{2\gamma}\xi$
where the deterministic force $F[\psi]$ is split into a conservative part (derivable from a free energy $H$ ) and a non-conservative part. The non-conservative part contains the non-Hermitian terms.
Path-Integral Formulation: They construct a generating functional $Z[J, \tilde{J}]$ using a response field $\tilde{\psi}$ . The dynamic action $A[\psi, \tilde{\psi}]$ is analyzed for invariance under time-reversal transformations.
Small-Noise Expansion: The system is expanded around a stationary deterministic solution $\psi_0$ $ψ_{0}$ (where $F[\psi_0]=0$ $F [ψ_{0}] = 0$ ). The field is decomposed as $\psi = \psi_0 + \sqrt{\gamma}\psi_1$ $ψ = ψ_{0} + γ ψ_{1}$ .
- This linearizes the theory, allowing the calculation of correlation ( $C$ ) and response ( $G$ ) functions to leading order in noise amplitude $\gamma$ .
- The fluctuation operator $\mathcal{A}_0$ (linearized Langevin operator) is analyzed. Crucially, the authors allow $\mathcal{A}_0$ to have complex eigenvalues, distinguishing Hermitian (real eigenvalues) from non-Hermitian (complex eigenvalues) cases.
Harada-Sasa Relation: The authors derive a generalized Harada-Sasa relation connecting the local Entropy Production Rate (LEPR), $\sigma_\psi$ , to the FDT violation difference, $\Delta$ .

3. Key Contributions and Theoretical Results

A. Characterization of TRS Breaking via FDT

The authors demonstrate that the violation of the Fluctuation-Dissipation Theorem is a direct signature of non-Hermiticity:

FDT Violation ( $\Delta$ ): In the frequency-momentum domain, the FDT difference $\Delta(\omega, q)$ is non-zero if and only if the fluctuation operator has non-vanishing imaginary eigenvalues ( $\text{Im}(\lambda_q) \neq 0$ ).
Scaling: The FDT violation scales linearly with the non-Hermitian component (denoted as $\zeta_q$ , the imaginary part of the eigenvalue).
$\Delta \propto \zeta_q$
This implies FDT violations are a highly sensitive, first-order probe of non-equilibrium driving.

B. Entropy Production Rate (EPR) Scaling

A central theoretical finding is the specific scaling of the Entropy Production Rate with non-Hermiticity:

Quadratic Dependence: Unlike the FDT violation, the local EPR ( $\sigma_\psi$ ) scales quadratically with the non-Hermitian component:
$\sigma_\psi \propto \zeta_q^2$
Physical Interpretation: This quadratic dependence ensures that the EPR is always non-negative (satisfying the second law of thermodynamics), regardless of the sign of the non-Hermitian term (i.e., regardless of the direction of the "drive").
General Formula: The authors derive a closed-form expression for the local EPR directly from the Langevin equation for arbitrary stationary states (uniform or non-uniform):
$\sigma_\psi(x) = \text{diag}_x \left[ \kappa^{-1} \mathcal{F}_a \mathcal{C} \mathcal{F}_a^\dagger (\kappa^{-1})^\dagger \right]$
where $\mathcal{F}_a$ is the anti-symmetric (anti-Hermitian) part of the dynamics, $\mathcal{C}$ is the covariance matrix, and $\kappa$ relates to the noise kernel.

C. Extension of Li and Cates Formalism

The work extends the methodology of Li and Cates (originally for Hermitian systems) to non-Hermitian operators. They show that even if the noise kernel and fluctuation operator commute (a condition where EPR would vanish in Hermitian systems), a non-zero EPR persists if the operator is non-Hermitian.

4. Application: Non-Reciprocal Model A

To illustrate the framework, the authors apply it to a minimal non-Hermitian extension of the Ginzburg-Landau $\psi^4$ theory, modeling a non-reciprocal Ising model with "vision cone" interactions.

The Model: The deterministic force includes a self-advection term: $\Gamma F[\psi] \supset \lambda \psi (\mathbf{v} \cdot \nabla)\psi$ . This term is anti-Hermitian.
Uniform Phase:
- In the disordered phase ( $\psi_0 = 0$ ), the EPR vanishes.
- In the ordered phase ( $\psi_0 = \pm m$ ), the EPR is non-zero and proportional to $m^2 \lambda^2$ .
Domain Wall (Non-Uniform Phase):
- The authors calculate the EPR across a domain wall (interface between ordered phases).
- Localization: The entropy production is localized at the interfaces.
- Behavior: The EPR vanishes at the center of the domain wall (where the order parameter $\psi_0$ is zero and symmetry is restored) and in the bulk uniform phases. It exhibits a double-peak structure near the walls of the interface.
- Orientation Dependence: The magnitude of the EPR depends on the angle between the vision cone vector $\mathbf{v}$ and the domain wall normal. The EPR is maximized when $\mathbf{v}$ is orthogonal to the wall.

5. Significance and Conclusions

Unified Characterization: The paper provides a rigorous link between microscopic non-Hermitian forces and macroscopic irreversibility. It establishes that while FDT violations are linear probes of non-equilibrium, the EPR is a quadratic, scalar measure.
Interface Localization: The analytical result that entropy production localizes at interfaces in non-uniform states supports previous numerical observations in active matter (e.g., MIPS), providing a theoretical explanation for why irreversibility concentrates at boundaries.
Generalizability: The framework is applicable to a broad class of non-equilibrium field theories, including those arising in active matter, dissipative quantum systems, and non-reciprocal interactions.
Future Directions: The authors suggest extending this analysis to conserved dynamics (Model B) and systems with hydrodynamic couplings to understand how non-reciprocity shapes spatial distributions of entropy production in more complex scenarios.

In summary, this work establishes a systematic, analytical toolkit for quantifying irreversibility in non-Hermitian field theories, revealing that the entropy production rate is fundamentally driven by the anti-Hermitian component of the linearized dynamics and is strictly localized at spatial interfaces in phase-separated states.