Uniform-in-temperature locality estimates for weakly interacting quantum systems

This paper establishes that weakly interacting quantum Hamiltonians exhibit uniform-in-temperature exponential decay of correlations and local indistinguishability by employing a low-temperature cluster expansion combined with a quantum probabilistic swapping trick.

The Gist

Imagine you have a giant, complex machine made of millions of tiny, interacting gears. In the world of quantum physics, this machine is a "many-body system," and the gears are atoms or particles. When this machine is hot, the gears jiggle wildly and interact chaotically. When it is cold, they settle down, but they still "talk" to each other.

The big question this paper asks is: If you look at just one small part of this machine, does it matter what the rest of the machine is doing?

Usually, in physics, we expect that if two parts of a system are far apart, they stop influencing each other. This is called locality. It's like sitting in a crowded room: if you are far away from someone shouting, you eventually stop hearing them.

However, there's a catch. Most mathematical tools used to prove that these distant parts don't influence each other break down when the machine gets very cold. It's as if the math only works when the room is warm, but fails when the room freezes. This is a problem because many modern technologies (like quantum computers) operate at extremely low temperatures.

The Core Discovery

The authors of this paper have found a way to prove that for a specific class of "weakly interacting" quantum machines, locality holds true even when the temperature drops to absolute zero.

They proved two main things:

1. Correlations Decay (The "Whisper" Effect): If you measure two distant parts of the system, the connection between them (correlation) fades away exponentially fast as the distance increases. Imagine a whisper: if you whisper to a friend, the person standing next to them hears it clearly, but the person across the room hears nothing. The authors proved that even in the freezing cold, this "whisper" dies out quickly over distance.
2. Local Indistinguishability (The "Blind Spot" Effect): This is the stronger result. It means that if you want to know what's happening in a small room (a local region), you don't need to know the state of the entire building. You can pretend the building ends right outside your door, and your calculations will be almost perfect. The "global" temperature of the whole system is indistinguishable from the "local" temperature of just your room, even in the deep freeze.

How They Did It: The "Swapping Trick"

To prove this, the authors used a clever mathematical strategy involving two main ingredients:

• Low-Temperature Clustering: They broke the complex system down into small, manageable "clusters" of interacting particles, similar to how you might break a large puzzle into smaller sections to solve it.
• The Swapping Trick: This is the star of the show. Imagine you have two different ways of arranging a deck of cards (representing the quantum states). The authors developed a method to "swap" parts of these arrangements. They showed that if two distant parts of the system are not connected in a specific way, you can swap the middle sections of the arrangements without changing the final outcome.

Think of it like this: If you have two long chains of people holding hands, and you want to know if the person at the very end of Chain A is holding hands with the person at the end of Chain B, you can prove they aren't by showing that you can swap the middle sections of the chains and the result looks exactly the same. If the swap works perfectly, it proves the two ends were never actually connected in the first place.

Why This Matters (According to the Paper)

The paper emphasizes that this result is robust and doesn't rely on the system being perfectly ordered (like a crystal). It works even if the system is "disordered" (like a messy pile of gears).

The authors highlight three specific applications where this "uniform" (temperature-independent) proof is useful:

1. Efficient Simulation: It allows scientists to simulate these quantum systems on classical computers much more easily, because they only need to look at small local pieces rather than the whole universe.
2. Thermal State Preparation: It helps in figuring out how to prepare these cold quantum states on quantum devices.
3. Response Theory: It lays the groundwork for understanding how these systems react to changes (like a slight push) at low temperatures, which is crucial for developing new quantum technologies.

The Bottom Line

Before this paper, we knew quantum systems were "local" (distant parts don't affect each other) at high temperatures, but we weren't sure if this held up in the deep freeze. This paper says: Yes, for a wide class of weakly interacting systems, the "locality" rule is unbreakable, whether the system is hot or cold. They achieved this by inventing a new mathematical "swapping trick" that works perfectly even when the temperature is near absolute zero.

Technical Summary

Technical Summary: Uniform-in-Temperature Locality Estimates for Weakly Interacting Quantum Systems

Problem Statement
The locality of thermal (Gibbs) states is a fundamental property in quantum many-body physics, essential for applications ranging from thermalization and response theory to the efficient classical and quantum simulation of quantum systems. Locality is typically characterized by two properties:

1. Decay of Correlations (DoC): The exponential decay of the covariance between local observables supported on spatially separated regions.
2. Local Indistinguishability (LI): The ability to approximate the expectation value of a local observable in a global thermal state using a local thermal state defined on a sufficiently large subregion.

While these properties are well-understood in high-temperature regimes (via cluster expansions) and for gapped ground states at zero temperature, establishing them uniformly in temperature—particularly in the low-temperature limit ($\beta \to \infty$)—remains a significant challenge. Existing techniques often yield correlation lengths that diverge as temperature decreases, or they rely on conditions (such as translation invariance) that limit their applicability. The central problem addressed is to identify conditions under which DoC and LI hold with correlation lengths that are uniformly bounded in $\beta$ for quantum systems in arbitrary spatial dimensions.

Methodology
The authors develop a new analytical framework that combines a low-temperature cluster expansion with a quantum version of a probabilistic swapping trick.

1. Model Setup: The study focuses on lattice systems on $\Lambda \subset \mathbb{Z}^D$ with finite-dimensional on-site Hilbert spaces. The Hamiltonian is $H = H_0 + V$, where:

   • $H_0 = \sum h_x$ consists of on-site terms with a unique ground state and a spectral gap of at least 1.
   • $V = \sum v_x$ represents weak, finite-range interactions.
   • Crucially, the interaction terms satisfy a relative form bound: $|\langle \psi, v_x \psi \rangle| \leq a \langle \psi, H_0 \psi \rangle$ for a small $a \in (0,1)$. This ensures the perturbed spectrum remains gapped and the ground state remains a product state, effectively excluding phase transitions at low temperatures.

2. Cluster Expansion and Analyticity: The proof begins by rewriting the exponential of the Hamiltonian using an inclusion-exclusion argument (following [67]). This decomposes the Gibbs state into a sum over configurations (subsets of the lattice). The authors exploit the analyticity of the interaction terms to bound the norms of the resulting expansion terms.

3. The Swapping Trick: A core innovation is the adaptation of a probabilistic "swapping trick" (originally developed for lattice gauge theories in [1]) to the quantum setting.

   • The truncated correlation function (or the difference in local expectations) is expressed as a difference between products of cluster expansions.
   • The authors define superclusters (maximally connected components formed by the union of configurations from different terms).
   • They demonstrate an exact algebraic equality: terms where the supports of observables lie in disjoint superclusters cancel out exactly between the two products.
   • Consequently, the non-vanishing contribution comes only from configurations where the observables lie within the same supercluster.

4. Combinatorial Estimates: The remaining sum is restricted to superclusters connecting the supports of the observables. Using percolation-type estimates (Lemma 3.12), the authors bound the number of such connected sets and show that the sum converges exponentially with the distance between observables, provided the interaction strength $a$ is sufficiently small.

Key Contributions and Results
The paper establishes the following main results for the class of weakly interacting Hamiltonians described above:

• Uniform Decay of Correlations (Theorem 2.2): For any inverse temperature $\beta \in (0, \infty)$, the system exhibits exponential decay of correlations:
  $$|\text{Tr}(AB\rho_\Lambda) - \text{Tr}(A\rho_\Lambda)\text{Tr}(B\rho_\Lambda)| \leq C_1 e^{C_2(|X|+|Y|)} \|A\|\|B\| e^{-d(X,Y)/\xi}$$
  Crucially, the correlation length $\xi$ is independent of $\beta$.

• Uniform Local Indistinguishability (Theorem 2.3): The difference between the expectation value of a local observable $B$ in the global state and a local state $\rho_{\Lambda'}$ decays exponentially with the distance to the boundary of $\Lambda'$:
  $$|\text{Tr}(B\rho_\Lambda) - \text{Tr}(B\rho_{\Lambda'})| \leq C_1 e^{C_2|Y|} \|B\| e^{-d(Y, \Lambda \setminus \Lambda')/\xi_{LI}}$$
  Like DoC, the decay rate $\xi_{LI}$ is uniform in temperature.

• Thermodynamic Limit (Corollary 2.5): The authors prove that the finite-volume Gibbs states converge to a unique thermodynamic limit state that satisfies both DoC and LI uniformly.

• Robustness: Unlike previous approaches relying on Pirogov-Sinai theory (which typically require translation invariance and multiple ground states), this method works for disordered Hamiltonians and does not assume translation invariance. It also improves upon existing low-temperature DoC bounds in one dimension by providing a uniform correlation length.

• Local Perturbations Perturb Locally (LPPL): The paper also addresses the LPPL principle (Theorem 4.1). However, the authors note that their current method yields a bound that deteriorates exponentially as $\beta \to \infty$ (scaling as $e^{2\beta\|W\|}$), and they conjecture that a uniform bound holds but remains unproven by this specific technique.

Significance and Claims
The paper claims to provide the first proof of Local Indistinguishability (LI) that is uniform in temperature for quantum systems in arbitrary dimensions.

• Conceptual Simplicity: The authors argue that their approach, utilizing the swapping trick, is conceptually simpler and more self-contained than the complex quantum versions of Pirogov-Sinai theory used in prior works [13, 22, 23].
• Overcoming Low-Temperature Divergence: By avoiding tools like quantum belief propagation (which introduce constants diverging as $T \to 0$), the authors achieve bounds where the correlation length does not blow up at low temperatures.
• Applications: The results imply that the "locality of temperature" is a robust property even at low temperatures. This has direct implications for:
  • The efficient classical simulability of thermal states.
  • The efficient preparation of thermal states on quantum devices.
  • The development of robust response theories for low-temperature interacting systems (though the detailed application to response theory is deferred to future work [40]).

The authors emphasize that while the results are restricted to weakly interacting systems (satisfying the form bound), the method's ability to handle disorder and lack of translation invariance represents a significant step forward in understanding the locality of quantum thermal states.