Convexity and non-Markovianity of Weyl Maps

This paper establishes a complete algebraic classification of Weyl dynamical maps on finite-dimensional systems using the Hermite normal form, revealing that non-Markovianity is non-additive under convex mixing and demonstrating the existence of irreducible eternally non-Markovian maps in dimensions higher than qubits, thereby extending the theory of quantum memory effects beyond the Pauli framework.

The Gist

Imagine a quantum system (like a tiny computer chip) as a dancer trying to perform a routine. Usually, the dancer is surrounded by a noisy crowd (the environment). If the crowd's noise is random and forgets the dancer instantly, the dancer's performance is "Markovian"—it's smooth, predictable, and has no memory of past mistakes.

However, sometimes the crowd remembers the dancer's previous steps and reacts to them later. This creates "Non-Markovian" dynamics, where the system has memory. This memory can be a bug (causing errors) or a feature (helping with complex tasks).

This paper explores a specific type of quantum dancer called a Weyl Map. While most previous studies only looked at simple 2-step dancers (qubits), this paper investigates dancers with more steps (higher dimensions, or "qudits"). The authors use a mathematical tool called the Hermite Normal Form to organize the possible moves into neat groups, much like sorting a deck of cards by suit and rank.

Here are the main discoveries, explained through simple analogies:

1. The "Uniformity" Rule for Smooth Dancing

The paper first asks: When does a single dancer perform a perfectly smooth, memoryless routine (a "semigroup")?

• The Finding: If the dancer uses a mix of moves where some moves are used more often than others (non-uniform), they cannot perform a smooth, memoryless routine. It's like trying to drive a car where you randomly press the gas and brake with different intensities; you can't maintain a steady speed.
• The Exception: The routine is smooth only if the dancer uses all their available moves with equal weight (isotropic). If they do this, they can perform a perfect, memoryless dance.

2. The "Mixing" Magic: Erasing Memory

One of the most surprising findings is about what happens when you mix different dancers together.

• The Scenario: Imagine you have several dancers, each of whom is terrible at forgetting. They are "eternally non-Markovian," meaning they hold onto memories of every step forever.
• The Magic: The authors prove that if you mix these "forgetful" dancers together in a specific way, the resulting group dance can become perfectly memoryless.
• The Analogy: It's like taking several people who are terrible at keeping a secret (they always talk about the past) and having them all talk at once. The noise cancels out, and suddenly, the group seems to have no memory of anything. This shows that memory is not additive; mixing bad memory can sometimes create good memory (or rather, no memory).

3. The "Irreducible" Memory (A New Discovery)

In the old world of simple 2-step dancers (qubits), you needed to mix two different types of bad dancers to create a "eternal memory" effect. You couldn't get it from just one.

• The New Discovery: In these higher-dimensional dancers (Weyl maps), the authors found "irreducible" eternal memory. This means a single, individual dancer can naturally hold onto memories forever without needing to be mixed with anyone else.
• The Analogy: In the old days, you needed a committee of people to remember a secret forever. Now, the authors found that a single person can be a "super-rememberer" on their own. This is a unique trait of higher-dimensional systems that doesn't exist in the simpler 2-step world.

4. The "Crowd Control" Limit

The paper also asks: How many different memory-holding dances can we mix together before the memory disappears?

• The Finding: There is a limit to how many distinct "memory groups" you can mix before the system becomes memoryless.
• The Analogy: Imagine you have a room full of people, each remembering a different secret. If you mix too many groups together, the secrets get diluted, and the room becomes "forgetful." The paper calculates exactly how many groups you can mix before you hit that "forgetting" point. Interestingly, in these higher-dimensional systems, you can mix many more groups than in the simple 2-step systems before you lose the memory effect.

Summary

The paper builds a bridge between the geometry of a "discrete phase space" (a mathematical grid of possible moves) and the behavior of quantum memory.

• Uniformity creates smooth, memoryless motion.
• Mixing can either erase memory (turning eternal memory into nothing) or create eternal memory (turning smooth motion into a memory-holding one), depending on the specific mathematical structure of the groups involved.
• Higher dimensions allow for "super-rememberers" that exist on their own, a phenomenon impossible in simpler systems.

The authors use a specific example of a 3-step dancer (a qutrit) to show how these transitions happen, proving that the rules of quantum memory change significantly when you move beyond the simplest systems.

Technical Summary

Technical Summary: Convexity and non-Markovianity of Weyl Maps

Problem Statement
The dynamics of open quantum systems are traditionally modeled using Markovian approximations, where environmental correlations decay rapidly, leading to memoryless evolution described by quantum dynamical semigroups (GKSL structure). However, realistic systems often exhibit non-Markovian behavior characterized by memory effects. While the convex structure of quantum dynamical maps—specifically how mixing Markovian channels can generate non-Markovianity—has been extensively studied in qubit systems (Pauli maps), these results have not been fully generalized to higher-dimensional systems (qudits). Existing literature lacks a comprehensive understanding of the algebraic structures, semigroup properties, and convex mixing mechanisms governing Weyl dynamical maps in finite-dimensional Hilbert spaces ($d > 2$).

Methodology
The authors employ a rigorous algebraic framework based on the discrete phase space $\mathbb{Z}_d \times \mathbb{Z}_d$ and the properties of Weyl operators. The methodology proceeds through several key steps:

1. Algebraic Classification: Utilizing the Hermite Normal Form (HNF), the authors provide a complete classification of the subgroups of the discrete phase space $\mathbb{Z}_d \times \mathbb{Z}_d$. This allows for a non-redundant categorization of the support of probability distributions defining Weyl maps into three types: cyclic subgroups, split rank-2 subgroups, and non-split rank-2 subgroups.
2. Spectral Analysis: The study analyzes the eigenvalues of Weyl dynamical maps and their generators. By relating the eigenvalues to the symplectic inner product and the dual subgroups ($G^\perp$), the authors derive explicit expressions for decay rates $\gamma_\alpha(t)$.
3. Divisibility and Semigroup Conditions: The paper investigates the conditions under which Weyl maps form quantum dynamical semigroups (time-independent generators) and satisfy CP-divisibility (Markovianity). It distinguishes between isotropic maps (uniform weight distributions) and anisotropic maps (non-uniform weights).
4. Convex Mixing Analysis: The authors examine the convex combinations of Weyl maps. They analyze two distinct scenarios: (a) mixing eternally non-Markovian (ENM) dephasing maps to see if they can generate Markovian semigroups, and (b) mixing distinct Markovian Weyl semigroups to determine conditions under which they generate ENM.
5. Explicit Examples: The theoretical results are illustrated using the qutrit case ($d=3$), where the discrete phase space structure is explicitly calculated to demonstrate transitions between Markovian, non-Markovian, and eternally non-Markovian regimes.

Key Contributions and Results

• Algebraic Framework: The paper establishes a complete classification of subgroups of $\mathbb{Z}_d \times \mathbb{Z}_d$ using HNF, providing a foundational algebraic tool for analyzing Weyl maps. It derives a formula for the total count of subgroups of a given order $K$ and identifies that the number of subgroups is maximized when $K=d$.
• Semigroup Properties:
  • Anisotropic Maps: It is proven that anisotropic Weyl maps with non-uniform weight distributions cannot form quantum dynamical semigroups if they possess at least two distinct nontrivial eigenvalues.
  • Isotropic Maps: Necessary and sufficient conditions are derived for isotropic Weyl maps to form Markovian semigroups. Specifically, the probability distribution function $p(t)$ must take the form $p(t) = \frac{|G|-1}{|G|}(1 - e^{-ct})$.
• Non-Additivity of Non-Markovianity:
  • Suppression of Memory: The authors prove that convex combinations of eternally non-Markovian Weyl dephasing maps can generate a Markovian semigroup. This demonstrates that non-Markovianity is not additive under mixing; memory effects can be suppressed by collective interference. This result depends on the parity of the cyclic subgroup order $\ell = d/s$.
  • Generation of Eternal Non-Markovianity: A general condition is established for when convex mixtures of $N$ distinct Weyl semigroups exhibit eternal non-Markovianity. The condition is $2 \le N < \min\left\{ \frac{d^2-1}{K-1}, N(K) \right\}$, where $N(K)$ is the total number of subgroups of order $K$.
• Irreducible Eternal Non-Markovianity: Unlike the qubit Pauli setting where ENM requires mixing, the paper identifies the existence of "irreducible" eternally non-Markovian Weyl dephasing maps in higher dimensions ($d \ge 3$). These are individual dynamical maps (generated by a single Weyl operator) that display eternal memory effects without any mixing mechanism.
• Generalization of Pauli Results: The study shows that generalized Pauli maps are a special subclass of Weyl maps. Consequently, the analysis of Weyl convex combinations encompasses and extends previous results on Pauli maps, revealing that higher-dimensional systems allow for larger mixtures of semigroups to sustain ENM compared to qubits.

Significance and Claims
The paper claims to uncover a fundamental connection among finite phase-space algebra, convex structures, and quantum memory effects. By extending the theory of non-Markovian dynamics beyond the Pauli framework, the work reveals that higher-dimensional systems possess intrinsically richer memory structures. Specifically, the discovery of irreducible ENM in individual Weyl maps highlights a qualitative distinction between qubit and qudit dynamics. The results demonstrate that the algebraic structure of the discrete Weyl phase space plays a governing role in memory effects, suggesting that finite phase-space methods are powerful tools for understanding high-dimensional quantum noise. The authors position this work as a step toward a deeper understanding of memory effects in high-dimensional open quantum systems, without proposing specific experimental implementations or immediate technological applications beyond the theoretical framework.