⚛️ quantum physics

Diameter truncated operator evolution

This paper introduces a diameter-based truncation method for simulating operator dynamics in out-of-equilibrium quantum systems, demonstrating through numerical studies of the kicked Ising and Heisenberg XXZ models that this physically motivated approach efficiently and accurately captures local correlation functions and transport properties by restricting simulations to operators supported on small lattice regions.

Original authors: Tom Holden-Dye, Max Marvell, Joel Mills, Christoper J. Turner, Arijeet Pal

Published 2026-06-29

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: Tom Holden-Dye, Max Marvell, Joel Mills, Christoper J. Turner, Arijeet Pal

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

Imagine you are trying to predict how a drop of ink spreads through a glass of water. In the quantum world, this "ink" is information, and the "water" is a complex system of particles. Usually, when you start with a simple, local piece of information (like a single atom spinning one way), it quickly gets messy. It spreads out, gets tangled with its neighbors, and becomes a giant, complicated web of connections.

Trying to track every single thread of this web is like trying to count every grain of sand on a beach while a hurricane is blowing. It's too much work for even the most powerful computers. This is the problem the authors of this paper are tackling.

The Old Way: Counting the "Heavy" Stuff

Previously, scientists tried to solve this by ignoring the "heavy" parts of the mess. They would say, "Let's only track the parts of the system that involve a few simple changes, and ignore the parts that involve many complex changes."

Think of it like cleaning a room. The old method was: "Ignore anything that weighs more than 5 pounds." This works okay, but it's still messy because a single heavy object might be huge and spread out, while a pile of tiny, light objects might be scattered everywhere.

The New Way: Measuring the "Size" of the Mess

The authors introduce a new method called Diameter-Truncated Operator Evolution (DTOE). Instead of weighing the mess, they measure its size (or "diameter").

Here is the analogy:
Imagine you are tracing a rumor spreading through a town.

The Old Method asked: "How many people are involved in this rumor?" If the rumor involves 100 people, it's too complex to track.
The New Method (DTOE) asks: "How far apart are the people involved?" If the rumor is spread across the whole town (a large diameter), we ignore it. But if the rumor is just happening in one small neighborhood (a small diameter), we track it carefully.

The authors argue that for many interesting things we want to measure (like how two specific points in the system talk to each other), the "long-distance" rumors don't actually matter. Only the "local neighborhood" rumors contribute to the answer. By cutting off anything that spreads too far, they can simulate the system much faster and more efficiently.

Why This Works: The "Easy" and "Hard" Zones

The paper explains that space and time in these quantum systems have two different "zones":

The "Easy" Zone: In this area, the information stays relatively local. It doesn't spread out wildly. Here, the new method works perfectly, like walking on a flat sidewalk.
The "Hard" Zone: In this area, information spreads out very fast and gets tangled. Here, the method has to work harder, like trying to walk through a dense forest.

The authors show that for many systems, the "Easy" zone is big enough that we can get accurate answers without needing to simulate the entire impossible "Hard" zone.

What They Tested

To prove their idea works, they tested it on two famous quantum models (think of these as two different types of "quantum machines"):

The Kicked Ising Model: A system that is chaotic and unpredictable. They found that even when the system was very chaotic, their "size-based" method could still predict how the system behaved for a long time, matching results that usually require supercomputers.
The Heisenberg XXZ Model: A system that has some rules it must follow (conservation laws). They used their method to measure how "heat" or "spin" moves through the system. They found it could accurately predict whether the movement was fast (ballistic), slow (diffusive), or somewhere in between, matching theoretical predictions perfectly.

The Bottom Line

The paper claims that by changing how we decide what to ignore—switching from "ignoring heavy things" to "ignoring things that are too big"—we can simulate complex quantum systems much more efficiently. They didn't use any supercomputers to get these results; they just used a smarter way of looking at the problem.

This method allows scientists to see further into the future of these quantum systems than before, accurately predicting how they behave without getting lost in the infinite complexity of the full system.

Technical Summary: Diameter-Truncated Operator Evolution (DTOE)

Problem Statement
Simulating the dynamics of out-of-equilibrium many-body quantum systems is fundamentally challenging due to the rapid growth of complexity. As these systems evolve, entanglement, non-stabilizerness, and non-Gaussianity increase, causing initially local operators to become highly non-local and "scrambled." This growth renders traditional tensor-network simulations intractable beyond very short timescales, as they cannot efficiently capture the full dynamical evolution of the wavefunction or all possible observables. While late-time thermalization allows for hydrodynamic descriptions, the approach to equilibrium at short and intermediate times remains difficult to compute accurately.

Methodology
The authors introduce Diameter-Truncated Operator Evolution (DTOE), a numerical method designed to simulate operator dynamics by restricting the simulation to a subspace of operators with a bounded spatial extent.

Core Concept: Unlike previous methods such as Dissipation-Assisted Operator Evolution (DAOE) or Pauli Propagation, which truncate operators based on their weight (the number of non-trivial Pauli terms in a string expansion), DTOE truncates based on diameter. The diameter is defined as the size of the real-space region on the lattice where the operator is non-trivially supported.
Implementation: The method is formulated within the framework of 1+1D brickwork quantum circuits. It utilizes a "folded" notation where operators are vectorized into a doubled Hilbert space. The time evolution is computed by applying transfer matrices that map intervals of diameter- $d$ operators.
Truncation Scheme: A hard truncation is employed: all operators with a diameter exceeding a fixed cutoff $d$ are strictly omitted from the simulation. The correlation functions are then reconstructed by chaining these transfer matrices to generate paths between local operators.
Theoretical Motivation: The approach relies on the observation that while the majority of operator norm flows into highly non-local trajectories, local two-point correlation functions are often dominated by "atypical" trajectories that remain within a bounded, constant-diameter region. For generic circuits, there exists an "easy" phase in spacetime (typically above a critical velocity $v_c$ ) where correlations are dominated by operators of constant $O(1)$ diameter. In this regime, DTOE offers a polynomial time-complexity ( $O(t^2)$ ) with constant memory scaling $O(t)$ .

Key Contributions and Results
The paper demonstrates the efficacy of DTOE through extensive numerical simulations on both integrable and non-integrable systems, without the use of high-performance computing resources.

Perturbed Dual-Unitary Circuits (Kicked Ising Model):
- The authors simulated the kicked Ising model with parameters $J=B=\pi/4+\eta$ and $h=1.2$ , varying the perturbation strength $\eta \in [0, 0.5]$ .
- They computed infinite-temperature two-point correlation functions ( $C_{XX}$ ) up to $t=100$ circuit layers.
- Results: For $\eta = 0.24588$ , low-diameter approximations ( $d=1, 2$ ) diverged from exact data, but higher diameters ( $d \ge 9$ ) achieved relative errors on the order of $10^{-2}$ to $10^{-3}$ . The method remained effective even in a highly non-perturbative regime, outperforming current digital quantum computer simulations which typically manage $\sim 25$ Floquet cycles.
Heisenberg XXZ Model:
- The method was applied to XXZ circuits to extract hydrodynamic transport exponents ( $\alpha$ ) for the scaling of correlation width $\sigma(t) \sim t^\alpha$ .
- Ballistic Regime ( $|\Delta| < 1$ ): DTOE with $d=10$ accurately extracted $\alpha = 1.00 \pm 0.01$ . The authors observed that the timescale $t^*$ over which the method remains accurate before drifting toward diffusion scales exponentially with the diameter $d$ ( $t^* \sim O(\exp(d))$ ).
- Superdiffusive Regime ( $|\Delta| = 1$ ): At the isotropic point, the method captured the theoretically predicted exponent $\alpha = 2/3$ (specifically $\alpha \in [0.6663, 0.6684]$ for $d=10$ ), despite the known difficulty of capturing long quasiparticle strings in this regime.
- Diffusive Regime ( $|\Delta| > 1$ ): The method accurately reproduced the diffusive exponent $\alpha = 1/2$ even at low diameters, as transport is dominated by short quasiparticle strings in this regime.

Significance and Claims
The paper claims that DTOE provides a "leaner" truncation protocol compared to weight-based methods. Theoretical analysis suggests that for a lattice of size $L$ , the number of operators with diameter $d$ scales as $O(L)$ , whereas the number of operators with weight $d$ scales as $O(L^d)$ . Consequently, for large $L$ , the ratio of retained operators in DTOE versus weight-based truncation vanishes, suggesting DTOE is intrinsically more efficient.

The authors assert that DTOE is physically well-motivated for systems where local correlations are dominated by bounded-diameter trajectories. They highlight that the method successfully extracts transport properties and correlation functions at timescales beyond current quantum hardware capabilities. Furthermore, they note that while their results are presented in 1+1D, existing literature suggests that in 2D generic circuits, the entire spacetime lattice may fall into the "easy" phase, implying DTOE could be particularly powerful in higher dimensions, though this generalization is not explicitly tested in this work. The authors also note that while DAOE is framed as an MPO method with a specific dissipation superoperator, framing DTOE similarly remains a challenge for future work.