🔢 mathematics

Quantum complexity and generalized area law in fully connected models

This paper establishes a generalized area law for gapped ground states of fully connected (geometrically non-local) Hamiltonians, demonstrating that their bipartite entanglement entropy grows at most logarithmically with system size, thereby enabling efficient matrix product state approximations and polylogarithmic-time computation for permutation-invariant cases.

Original authors: Donghoon Kim, Tomotaka Kuwahara

Published 2026-09-25

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Donghoon Kim, Tomotaka Kuwahara

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

In the vast landscape of quantum physics, scientists are constantly trying to understand how the tiniest building blocks of nature organize themselves. When many particles come together to form a complex system, they do not simply act as a collection of independent individuals; instead, they become deeply linked in a phenomenon called entanglement. This connection is so profound that the state of one particle instantly reflects the state of another, no matter how far apart they are. For decades, researchers have relied on a guiding principle known as the "area law" to make sense of this complexity. This rule suggests that in most ordinary materials, the amount of entanglement between two parts of a system depends only on the surface area where they touch, rather than the total volume of the material. This principle is crucial because it implies that even the most complex quantum states can be described using a manageable amount of information, allowing scientists to simulate them on computers. However, this rule was rigorously proven only for systems arranged in a single line. When scientists looked at systems where every particle interacts with every other particle simultaneously—ignoring physical distance entirely—the rules seemed to break down, and the complexity appeared to explode.

A team of researchers at RIKEN in Japan has now shown that this explosion of complexity is not inevitable, even in these extreme, fully connected systems. They focused on a specific type of quantum model where every component interacts with every other component, a setup often used to describe everything from the behavior of atomic nuclei to the physics of black holes. In these models, there is no geometric distance to separate particles; every site is effectively a neighbor to every other site. Intuitively, one might expect such a system to be hopelessly tangled, with entanglement spreading everywhere like a net. Yet, the researchers discovered that if the system has a stable energy gap—a buffer that keeps the ground state distinct from excited states—the entanglement remains surprisingly tame. They proved that the entanglement between any two groups of particles grows only very slowly as the system gets larger, specifically increasing at a logarithmic rate. This means that despite the lack of physical boundaries, the system naturally organizes itself so that each particle is only weakly connected to the rest, effectively restricting the chaos to a manageable level.

The key to this discovery lies in how the researchers analyzed the energy of the system. They assumed that the energy associated with any single particle remains bounded and does not grow uncontrollably as the system gets bigger. Under this condition, they found that the ground state of the system behaves almost like a collection of independent particles, with only a tiny fraction of the system deviating from this simple, orderly state. They demonstrated that configurations where many particles fluctuate wildly are so rare that they can be ignored for practical purposes. This insight allowed them to show that the entire complex quantum state can be approximated with high precision using a mathematical structure known as a matrix product state. In simpler terms, this means that even though the system is fully connected, it can be described by a computer algorithm that requires only a polynomial amount of memory relative to the system size, making it computationally feasible to study.

The findings become even more powerful when the system possesses a special kind of symmetry, where the rules do not change no matter how the particles are labeled. In this permutation-invariant setting, the researchers proved that the entanglement does not grow at all; it stays constant regardless of how large the system becomes. Furthermore, they developed a classical algorithm that can calculate the energy and structure of these ground states in a time that grows only with the logarithm of the system size. This is a dramatic improvement over previous methods, which would take exponentially longer as the system grew. The work suggests that the ability to efficiently describe and compute these complex states does not strictly depend on particles being close to one another in space, but rather on the underlying stability of their energy levels.

This result challenges the long-held view that geometric locality—the fact that things are only close to their immediate neighbors—is the only way to keep quantum complexity in check. By proving that a generalized area law holds for fully connected systems, the researchers have expanded the conceptual scope of what is possible in quantum physics. They showed that even in a world where every particle talks to every other particle, the ground state can remain simple and structured. This does not mean that all such systems are simple; the researchers noted that if the energy conditions are not met, such as in certain disordered models, the entanglement can indeed become overwhelming. However, for the broad class of systems with stable, bounded energy scales, the complexity is under control. This provides a new theoretical foundation for understanding a wide range of physical phenomena, from spin glasses to quantum simulators, and opens the door to more efficient ways of solving some of the most difficult problems in quantum many-body physics.

Technical Summary: Quantum Complexity and Generalized Area Law in Fully Connected Models

Problem Statement
The entanglement area law posits that the entanglement entropy of a gapped ground state scales with the boundary area of a subsystem rather than its volume. While rigorously established for one-dimensional (1D) local systems, the status of this law in higher-dimensional local systems remains unresolved, and it is generally known to fail in geometrically non-local systems. Fully connected (all-to-all interacting) systems represent an extreme case of non-locality where spatial dimensions lose relevance. In such systems, the entanglement structure is highly variable: some models (e.g., Lipkin-Meshkov-Glick) exhibit logarithmic scaling, while others (e.g., Sachdev-Ye-Kitaev or transverse-field Sherrington-Kirkpatrick) often display volume-law scaling. The central challenge is to determine under what structural conditions a generalized area law holds for gapped ground states in fully connected systems, despite the absence of geometric locality and the presence of extensive "boundaries" in any bipartition.

Methodology
The authors establish a generalized area law for gapped ground states of fully connected Hamiltonians under specific structural assumptions. The framework considers a system of $n$ qudits with a $k$ -local Hamiltonian $H$ . The key structural assumptions are:

Spectral Gap: The Hamiltonian possesses a constant spectral gap $\Delta$ independent of system size $n$ .
Bounded Local Energy Scale: The interaction strengths are scaled such that the energy associated with a single site remains bounded. Specifically, the sum of interaction norms involving any site $i$ is bounded by a constant $\bar{g}$ , achieved via Kac-type normalization (e.g., $1/n$ scaling for pairwise interactions). This distinguishes the work from models where local energy scales grow with $n$ (such as certain disordered SK or SYK models).

The proof strategy departs from traditional area law techniques (like Approximate Ground State Projection or Lieb-Robinson bounds) which rely on geometric separation. Instead, the authors employ a quantitative trade-off inequality linking the spectral gap to the suppression of fluctuations in local observables. The core logical flow involves:

Robustness and Bootstrap: Demonstrating that the ground state is robust under local unitary transformations that preserve the single-site reduced density matrix. By iteratively applying a "bootstrap" argument, they show that the spectral gap forces the weight of Schmidt coefficients outside the dominant local component to scale as $O(1/n)$ .
Emergent Local Product Structure: This weak entanglement per site implies that the ground state is locally concentrated on a "mean-field" state.
Concentration Bounds: The authors analyze the probability distribution of the total number of sites deviating from this mean-field configuration. They prove a stretched-exponential concentration bound, showing that the ground state has negligible weight on configurations with a large number of local deviations.
Dimension Reduction: This concentration allows the effective Hilbert space to be restricted to a low-dimensional subspace spanned by configurations with few deviations, facilitating efficient approximation.

Key Contributions and Results

Logarithmic Entanglement Bound:
For general fully connected gapped Hamiltonians satisfying the bounded local energy assumption, the bipartite entanglement entropy $S_{A,B}(|\Omega\rangle)$ grows at most logarithmically with system size:
$S_{A,B}(|\Omega\rangle) \leq C \left(\frac{\bar{g}}{\Delta}\right)^3 \log n$
This result holds despite the absence of geometric locality, demonstrating that low entanglement complexity does not strictly require geometric separation.
Efficient Matrix Product State (MPS) Approximation:
Due to the concentration of the ground state on configurations with few local deviations, the ground state admits an efficient MPS approximation. For any fixed accuracy $\epsilon$ , there exists an MPS $|\Omega_D\rangle$ with bond dimension:
$D = n^{O(\log^2(1/\epsilon))}$
For constant precision, the bond dimension scales polynomially with $n$ , enabling efficient classical representation.
Permutation-Invariant Systems:
In the specific case of permutation-invariant Hamiltonians, the authors prove a stronger result:
- Constant Entanglement: The entanglement entropy is bounded by a constant $O(1)$ , independent of system size.
- Polylogarithmic Time Algorithm: They construct a classical algorithm to compute the ground-state energy and an approximate ground-state vector with error $\epsilon$ in time $O(\text{polylog}(1/\epsilon) \log(n/\epsilon))$ . For inverse-polynomial precision ( $\epsilon = 1/\text{poly}(n)$ ), the complexity is polylogarithmic in $n$ . This is achieved by optimizing over a variational family of local bases within a symmetric subspace of dimension polynomial in the deviation number $z$ .

Significance and Claims
The paper claims to broaden the conceptual and computational scope of area laws by proving that low entanglement complexity can emerge in the extreme limit of infinite-dimensional (mean-field) systems without geometric locality. The results establish a provable area-law-type bound in a setting where traditional geometric arguments fail.

The authors explicitly distinguish their findings from counterexamples to generalized area laws (e.g., Ref. [40]) by emphasizing the necessity of the uniformly diluted interaction strength (Kac normalization). In their framework, the total interaction strength involving any fixed pair of sites decreases with $n$ , keeping the local energy scale bounded. This excludes mechanisms where local energy scales grow with system size, which can lead to volume-law entanglement.

Furthermore, the work addresses the Local Hamiltonian Problem (LHP) for gapped fully connected models. It implies that for constant precision, the problem lies in NP, as the polynomial-bond-dimension MPS serves as a classical witness verifiable in polynomial time. However, for inverse-polynomial precision, the bond dimension becomes quasi-polynomial, leaving the precise complexity class for that regime as an open question.

The authors note that while their logarithmic bound applies generally, the constant bound and polylogarithmic algorithm are specific to permutation-invariant systems. They identify the finite-pattern transverse-field quantum Hopfield model as a potential candidate for non-permutation-invariant systems where the general logarithmic bound applies, though identifying broader classes of such models remains an open problem. Finally, they suggest that extending these techniques to strongly long-range interacting systems with Kac normalization in finite-dimensional settings is a promising direction for future research.

Technical Summary: Quantum Complexity and Generalized Area Law in Fully Connected Models

More like this