Volume-Independent Spectral Stability of Energy-Truncated Effective Hamiltonians in Quantum Spin Systems

This paper establishes a volume-uniform spectral stability theorem for energy-truncated effective Hamiltonians in bounded finite-range quantum spin systems, proving that low-energy spectral subspaces remain stable with exponentially small errors in both finite and infinite volumes, thereby extending previous finite-volume results to the thermodynamic limit.

The Gist

Imagine you are trying to understand the behavior of a massive, complex machine made of billions of tiny, interacting gears (a quantum spin system). This machine is so big that it might be infinite in size. You are only interested in how the machine behaves when it is "quiet"—that is, in its lowest energy states.

However, calculating the exact behavior of every single gear is impossible. So, physicists use a trick: they build a simplified model (an "Effective Hamiltonian"). This model ignores the crazy, high-energy jitters of the gears and only focuses on the smooth, low-energy movements.

The big question is: Does this simplified model actually tell us the truth about the real machine?

The Problem: The "Size" Trap

In the past, scientists had a way to prove that the simplified model was accurate, but it only worked for small, finite machines. They tried to say, "The difference between the real machine and the model is tiny."

But here's the catch: as the machine gets bigger and bigger (approaching an infinite size), that "tiny difference" used to grow uncontrollably. It was like trying to measure the error of a map by looking at the whole world at once; the more land you added, the bigger the error got. This made it impossible to use the simplified model for truly infinite systems, which is what physicists really want to study.

The Solution: A New Way to Measure "Leakage"

This paper, by Ayumi Ukai, introduces a clever new way to measure the accuracy of the simplified model. Instead of trying to measure the direct "difference" between the two machines (which gets messy as the system grows), the author measures spectral leakage.

Think of the energy states of the machine as floors in a skyscraper:

• Low floors: The quiet, low-energy states we care about.
• High floors: The chaotic, high-energy states we ignore.

The simplified model is supposed to keep all its attention on the low floors. The "leakage" is how much of the simplified model's attention accidentally spills over onto the high floors of the real machine.

The author proves a surprising result: Even as the building gets infinitely tall, the amount of "leakage" stays small and controlled.

The Key Ingredients

To make this work, the author uses a few specific tools:

1. The "Cut-Off" (The Energy Limit): The simplified model is built by strictly cutting off any energy above a certain height (let's call it $M$). The paper shows that if you set this cut-off high enough, the "leakage" into the high-energy zones drops off exponentially. This means if you double the cut-off height, the error doesn't just get half as bad; it gets astronomically smaller.
2. Local Rules: The proof relies on the fact that the gears only interact with their immediate neighbors (finite-range interactions). Because the chaos is local, the size of the whole system doesn't matter. The error depends only on the local neighborhood and the cut-off height, not on how many total gears there are.
3. The "Spectral Overlap" Method: Instead of comparing the machines directly, the author compares the spaces they occupy. They prove that the "low-energy room" of the simplified model fits almost perfectly inside the "low-energy room" of the real machine, with very little of it sticking out into the high-energy zone.

The Results

• For Finite Systems (Small Machines): The paper confirms that the low-energy "notes" (eigenvalues) of the simplified model are almost exactly the same as the real machine. The error is so small it's practically zero, and this holds true regardless of how big the machine is.
• For Infinite Systems (The Big Picture): This is the breakthrough. The author extends this proof to infinite systems. Even though an infinite system doesn't have a single "lowest note" in the traditional sense, the paper proves that the simplified model still correctly captures the structure of the low-energy states. It works in the "thermodynamic limit" (the limit of infinite size).

The Bottom Line

The paper solves a long-standing problem in quantum physics. It shows that you can safely use simplified, energy-truncated models to understand the low-energy behavior of quantum spin systems, even when those systems are infinitely large.

The author essentially says: "Don't worry about the size of the system. If you cut off the high-energy noise at a high enough level, your simplified model will stay 'grounded' in the low-energy reality, no matter how big the universe of gears gets."

This provides a rigorous mathematical foundation for using these simplified models to study complex phenomena like phase transitions and topological states in materials, ensuring that the math holds up even in the infinite limit.

Technical Summary

Technical Summary: Volume-Independent Spectral Stability of Energy-Truncated Effective Hamiltonians in Quantum Spin Systems

Problem Statement
The paper addresses the construction of effective Hamiltonians for quantum spin systems by truncating high-energy components. While such constructions are standard tools for analyzing low-energy structures, existing formulations face a structural obstruction when attempting to extend them to the thermodynamic limit (infinite volume). Specifically, the foundational finite-volume operator-norm estimate established by Arad, Kuwahara, and Landau (AKL) [1], which bounds $\|(H - \bar{H})\bar{P}_{low}\|$, cannot be made volume-uniform. As demonstrated in Appendix A of this work, this operator-norm bound necessarily depends on the total system volume in general cases. Consequently, the direct application of finite-volume operator-norm formulations fails for infinite-volume systems where individual low-energy eigenvalues may not exist, and the stability statement itself must be reformulated to remain meaningful in the thermodynamic limit.

Methodology
The author replaces the direct operator-norm control with a spectral-overlap formulation. Instead of estimating the difference between the original Hamiltonian $H$ and the truncated Hamiltonian $\bar{H}$ on a low-energy subspace, the paper quantifies the "leakage" of the low-energy spectral subspace of $\bar{H}$ into the high-energy spectral subspace of $H$.

The core technical approach involves:

1. Spectral Overlap Estimation: The central quantity analyzed is the norm of the projection of the low-energy subspace of $\bar{H}$ onto the orthogonal complement of the low-energy subspace of $H$:
   $$\|(I - E_H(-\infty, q)) E_{\bar{H}}(-\infty, p]\|$$
   This measures how much the low-energy states of the effective Hamiltonian deviate from the low-energy states of the original system.
2. Locality and Imaginary-Time Evolution: The proofs rely on the assumption of bounded, finite-range interactions. The author utilizes imaginary-time evolution estimates (Hadamard-type estimates) to control the decay of off-diagonal terms. Crucially, the proof strategy (specifically Lemma 3.5 and its infinite-volume extension, Proposition 4.5) focuses on the boundary region interactions ($H_{\partial L}$ or $H_X$) rather than the full Hamiltonian. This ensures that the decay constants depend only on local interaction parameters and the boundary size, not the total volume.
3. GNS Representation for Infinite Systems: For infinite-volume systems, the construction is performed within the Gelfand-Naimark-Segal (GNS) representation of an infinite-volume ground state $\omega$. The Hamiltonian $H_\omega$ is generally unbounded, requiring careful handling of domains and analytic extensions of operator-valued functions to justify the spectral estimates.

Key Results

• Finite-Volume Spectral Overlap (Theorem 3.2):
  For a finite system $\Lambda$, the paper establishes a volume-uniform bound on the spectral leakage. For cutoffs $M$ and energy windows defined by $p, q$, the bound takes the form:
  $$\|(I - E_{H_\Lambda}(-\infty, q)) E_{\bar{H}_\Lambda}(-\infty, p]\| \leq \frac{p + 2\|H_{\partial L}\| + \delta(p,q)}{q + 2\|H_{\partial L}\|}$$
  where $\delta(p,q)$ contains a term exponentially decaying in the cutoff $M$. The constants depend on local interaction parameters and the boundary region but are independent of the total volume $|\Lambda|$.

• Finite-Volume Eigenvalue Stability (Theorem 3.4):
  The spectral overlap bound implies stability of low-lying eigenvalues. If $\epsilon_j$ and $\bar{\epsilon}_j$ are the $j$-th eigenvalues of $H_\Lambda$ and $\bar{H}_\Lambda$ respectively, then:
  $$\epsilon_j - 2\delta_j \leq \bar{\epsilon}_j \leq \epsilon_j$$
  The error term $\delta_j$ is exponentially small in the cutoff $M$ and volume-independent.

• Infinite-Volume Spectral Overlap (Theorem 4.3):
  The main result extends the spectral-overlap estimate to the GNS representation of an infinite-volume ground state. The theorem provides a volume-uniform bound on the leakage between the spectral subspaces of $H_\omega$ and $\bar{H}_\omega$. This formulation is applicable even when the spectrum is not discrete.

• Rank-Threshold and Ground State Consequences (Corollary 4.4):
  In the infinite-volume setting, the spectral overlap estimate yields consequences for rank thresholds (generalized eigenvalues). Specifically, if the ground state is locally unique with a spectral gap $\Delta$, the overlap between the true ground state vector $\Omega$ and the effective ground state vector $\bar{\Omega}$ satisfies:
  $$|\langle \Omega, \bar{\Omega} \rangle|^2 \geq 1 - \left(\frac{2\delta_0 + \eta}{\Delta}\right)^2$$
  This confirms the stability of the ground state projection in the thermodynamic limit.

Significance and Claims
The paper claims to extend and strengthen the effective-Hamiltonian mechanism of Arad, Kuwahara, and Landau [1] by reformulating it as a volume-uniform spectral-overlap estimate.

• Resolution of Obstruction: The work identifies that the operator-norm formulation is inherently too strong for a volume-uniform theory. By shifting to spectral overlap, the author bypasses the volume-dependent obstruction, allowing the effective Hamiltonian construction to be applied directly to infinite-volume systems.
• Thermodynamic Limit Applicability: Unlike previous finite-volume results, this formulation remains meaningful in the thermodynamic limit where individual eigenvalues may not exist, providing a rigorous justification for the analytic and domain issues arising from unbounded GNS Hamiltonians.
• Generality: The results apply to general low-energy spectral windows, not just the ground state, and cover bounded finite-range interactions without assuming translation invariance.

The author notes that while the spectral-overlap estimate is weaker than the operator-norm estimate in terms of direct operator comparison, it is sufficient for the spectral conclusions required (eigenvalue stability and ground state projection stability) and is the necessary formulation for volume-uniformity. The paper does not propose new experimental applications but positions this result as a foundational tool for future studies in phase classification, symmetry-protected topological phases, and low-energy approximations of quantum dynamics.