Solving Lévy Sachdev-Ye-Kitaev Model

This paper presents an exact large-$N$ solution of the Lévy Sachdev-Ye-Kitaev (LSYK) model, demonstrating how its tail exponent $\mu$ continuously interpolates between a free theory and the maximally chaotic Gaussian SYK model while exhibiting non-maximal chaos and distinct thermodynamic properties in the intermediate regime.

The Gist

Imagine you are trying to understand how a giant, complex machine works. In the world of physics, this machine is made of tiny particles called fermions (think of them as the "atoms" of a quantum computer). Usually, these particles interact in a very messy, random way.

For a long time, physicists have studied a specific model called the SYK model (named after Sachdev, Ye, and Kitaev). Think of the SYK model as a "perfectly messy" machine where every single particle talks to every other particle at the exact same time. This model is special because, despite the chaos, it's actually solvable (we can write down the math to predict what happens). It's also famous for being "maximally chaotic," meaning it scrambles information as fast as the laws of physics allow—like a drop of ink instantly turning a whole glass of water black.

However, real-world materials aren't always perfectly connected. Sometimes, connections are weak, or some particles don't talk to each other at all. This is called a "sparse" system. The problem is: Sparse systems are incredibly hard to solve mathematically.

The Big Idea: The "Fat-Tailed" Solution

This paper introduces a clever workaround. The authors propose a new version of the machine called the Levy SYK model.

Instead of the interactions between particles being "normal" (like a bell curve where most interactions are average and extreme ones are rare), they use a Levy distribution.

• The Analogy: Imagine you are rolling dice to decide how hard two particles bump into each other.
  • In the standard model, you roll a normal die. You usually get a 3 or 4. Getting a 100 is almost impossible.
  • In the Levy model, you roll a "magic die" with fat tails. Most of the time, you get small numbers, but occasionally, you roll a number so huge (like 1,000,000) that it dominates the whole game.

These "huge numbers" act like the strong connections in a sparse system. Even though the model technically connects everyone to everyone, the math treats the tiny connections as noise and focuses only on the few "giant" connections. This mimics a sparse system but keeps the math solvable!

The Control Knob: The Parameter $\mu$

The authors introduce a dial called $\mu$ (mu) that controls how "fat" those tails are.

• $\mu = 2$ (The Standard): The tails are thin. This is the original, perfectly connected, maximally chaotic SYK model. It's like a crowded party where everyone is shouting at everyone else.
• $\mu = 0$ (The Freeze): The tails are so fat that the "giant" numbers take over completely, but in a way that breaks the chaos. The system becomes "free" or "frozen." It's like a library where everyone is silent and not interacting at all.
• $0 < \mu < 2$ (The Sweet Spot): This is the new discovery. The system is in the middle. It's still chaotic (shuffling information), but not maximally chaotic. It's like a busy coffee shop: people are talking and moving, but not in the frantic, all-at-once chaos of the standard model.

What Did They Find?

The authors did two main things: they solved the math and checked the "temperature" of the system.

1. The Chaos Meter:
They measured how fast the system scrambles information (chaos).

• At $\mu=2$, it scrambles at the maximum speed allowed by nature.
• At $\mu=0$, it doesn't scramble at all.
• In between, the scrambling speed slows down smoothly. It's a continuous slide from "Super Fast" to "Stopped."

2. The Thermodynamics (Heat and Energy):
They calculated how the system behaves when you heat it up or cool it down.

• In standard physics, when you cool something down, it usually settles into a neat, predictable state.
• In this new Levy model, as you cool it down, the system gets stuck in a weird state where it has a huge number of possible configurations (high entropy) even at very low temperatures.
• The Black Hole Connection: In the world of string theory, these chaotic quantum systems are often thought to be the "holographic" description of black holes. The authors suggest that this new model describes a weird, "shrinking" black hole. As you change the dial $\mu$, the black hole's horizon (its edge) changes size in a strange way that depends on the temperature, unlike the standard black holes we usually study.

Why Does This Matter?

Think of this paper as finding a universal remote control for quantum chaos.

• Before, we had a remote with only two buttons: "Max Chaos" and "No Chaos."
• Now, we have a remote with a slider. We can dial in any level of chaos we want and still do the math.

This is huge because it gives physicists a tool to study "sparse" systems (which are more realistic for real materials and quantum computers) without losing the ability to solve the equations. It bridges the gap between the idealized, perfectly connected world of theory and the messy, disconnected world of reality.

In a nutshell: The authors found a mathematical "magic trick" using heavy-tailed probability distributions to create a tunable model of quantum chaos. This model acts like a bridge, allowing scientists to smoothly transition from a frozen, non-chaotic state to the most chaotic state possible, helping us understand how complex quantum systems behave when they aren't perfectly connected.

Technical Summary

1. Problem Statement

The Sachdev-Ye-Kitaev (SYK) model is a paradigmatic solvable model of quantum chaos and holography, typically defined by all-to-all interacting Majorana fermions with Gaussian-distributed random couplings. While the standard SYK model ($\mu=2$) is maximally chaotic and solvable in the large-$N$ limit, variants with sparse couplings (which mimic integrable-to-chaotic crossovers) generally lose this solvability.

The authors address the open problem of constructing a solvable model that interpolates between a free theory and the maximally chaotic Gaussian SYK model, effectively emulating the physics of sparse SYK without sacrificing analytical tractability. They propose and solve the Lévy Sachdev-Ye-Kitaev (LSYK) model, where the random interaction couplings are drawn from a Lévy-Stable distribution characterized by a tail exponent $\mu \in [0, 2]$.

• $\mu = 2$: Recovers the standard Gaussian SYK (maximally chaotic).
• $\mu = 0$: Corresponds to a free theory (integrable).
• $0 < \mu < 2$: Represents an intermediate regime with non-maximal chaos.

2. Methodology

The authors employ a combination of path integral techniques, bosonic oscillator representations, and large-$N$ saddle-point analysis to solve the model.

• Hamiltonian and Disorder: The Hamiltonian involves $q$-body interactions with couplings $J_I$ sampled from a Lévy-Stable distribution. Unlike Gaussian disorder, Lévy distributions have fat tails ($P(X) \sim |X|^{-1-\mu}$), leading to divergent variance for $\mu < 2$.
• Partition Function and Averaging: The disorder-averaged partition function is computed using the characteristic function of the Lévy distribution. A key technical hurdle is that the partition function diverges for real inverse temperature $\beta$ due to the heavy tails. The authors resolve this by performing a Wick rotation to imaginary temperature ($\beta = -ik$), averaging, and then rotating back.
• Bosonic Oscillator Representation: To handle the non-Gaussian exponent $\exp(-V(G)^\mu)$ in the action (where $V(G)$ is a collective mode), the authors introduce a novel bosonic oscillator representation. They map the non-linear term to an integral over bosonic fields using a displacement operator formalism. This allows the action to be cast in a form suitable for deriving Schwinger-Dyson (SD) equations.
• Saddle Point Analysis:
  • Static Saddles: An initial ansatz assuming site-independent bosonic modes leads to divergent contributions.
  • Fluctuating Saddles: To cure the divergence and capture the emergent sparsity, the authors introduce "frozen" fluctuations $\xi_I$ (independent of time but dependent on the interaction index $I$) subject to a global constraint $\sum \xi_I = 0$. This leads to a finite, well-defined on-shell action.
• Schwinger-Dyson Equations: From the effective action, they derive the SD equations for the Green's function $G$ and self-energy $\Sigma$. The self-energy contains a non-trivial, $\beta$-dependent prefactor arising from the Lévy statistics.

3. Key Contributions and Results

A. Analytical Solutions

• Large-$q$ Limit: The authors solve the SD equations analytically in the large-$q$ limit. They find that the Green's function retains a Liouville-type structure but with a parameter $\nu$ that depends non-trivially on both $\mu$ and $\beta$.
  • The scaling of $\nu$ changes with $\mu$: $\nu \sim \beta^{\mu/2}$ at high temperatures ($\beta \to 0$) and $\nu \to 1$ at low temperatures.
• Infrared (IR) Regime: In the strong coupling limit, the model exhibits emergent reparameterization symmetry (conformal invariance) similar to Gaussian SYK, but with a modified central charge and a shrinking conformal window. The UV cutoff for the conformal regime scales as $\xi \sim \beta^{(2-\mu)/(2+\mu)}$, indicating that the conformal window shrinks as $\mu$ decreases.
• Schwarzian Action: The effective action for reparameterizations retains the Schwarzian form $S \sim \int \{f, \tau\}$, but the coefficient (related to the specific heat) scales as $C \sim \beta^{(2-\mu)/(2+\mu)}$, differing from the linear scaling in Gaussian SYK.

B. Chaos Exponents

The paper quantifies quantum chaos using two distinct exponents:

1. Krylov Exponent ($2\alpha$): Derived from the growth of Lanczos coefficients in the Krylov basis. It is given by $2\alpha = 2\pi\nu/\beta$.
2. Lyapunov Exponent ($\lambda_L$): Extracted from the 4-point function (OTOC) via a Schrödinger-like eigenvalue problem.
   • Result: For Gaussian SYK ($\mu=2$), $\lambda_L = 2\alpha$, saturating the chaos bound (maximal chaos).
   • LSYK Finding: For $0 < \mu < 2$, $\lambda_L < 2\alpha$. The model exhibits non-maximal chaos. The ratio $\lambda_L / 2\alpha$ is determined by a factor $\gamma = \sqrt{-E_{min}} \leq 1$, which decreases as $\mu \to 0$.
   • Phase Diagram: The model transitions from free ($\mu=0$) to non-maximal chaotic ($0 < \mu < 2$) to maximally chaotic ($\mu=2$).

C. Thermodynamics

The authors compute thermodynamic quantities (entropy, free energy, specific heat) both analytically (large-$q$) and numerically (finite-$q$).

• Scaling Laws: The free energy and specific heat exhibit power-law scaling dependent on $\mu$. Specifically, the specific heat $C(T)$ scales as $T^{2\mu/(\mu+2)}$ at low temperatures, deviating from the linear $T$ scaling of the marginal Fermi liquid (Gaussian SYK).
• Non-Fermi Liquid Behavior: The system displays anomalous thermodynamics characteristic of non-Fermi liquids, with a ground state degeneracy that dominates at low temperatures for small $\mu$.
• Numerical Verification: Numerical solutions of the SD equations confirm the analytical large-$q$ results and the scaling behaviors of thermodynamic coefficients.

D. Holographic Dual (Bulk Interpretation)

The authors propose a holographic dual based on 2D dilaton gravity (Jackiw-Teitelboim gravity).

• The scaling of the Schwarzian coefficient implies a modification to the dilaton potential $V(\Phi) \sim \Phi^{(\mu+2)/(2\mu)}$.
• This leads to a black hole horizon radius scaling as $r_h \sim T^{2\mu/(\mu+2)}$.
• Physical Implication: For $\mu < 2$, the black hole horizon is "stiffer" and less sensitive to temperature changes compared to the standard SYK dual. As $\mu \to 0$, the horizon becomes frozen (temperature independent), corresponding to the integrable limit.

E. Alternative Description

In Appendix A, the authors provide an alternative derivation using the Stochastic Representation of Lévy variables. They show that the LSYK model can be viewed as an infinite sum of correlated Gaussian SYK models, establishing a non-trivial connection between the two frameworks.

4. Significance

• Solvable Sparse SYK: The LSYK model provides the first exactly solvable framework (in the large-$N$ limit) that captures the physics of sparse SYK models, bridging the gap between integrable and chaotic many-body systems.
• Controlled Chaos: It demonstrates that quantum chaos can be continuously tuned via the tail exponent $\mu$, revealing a regime of "non-maximal chaos" where the Lyapunov exponent is strictly less than the Krylov growth rate.
• Holography and Gravity: The results offer new insights into the AdS/CFT correspondence, suggesting that different types of disorder in the boundary theory correspond to specific modifications of the bulk dilaton potential and black hole thermodynamics.
• Methodological Advance: The introduction of the bosonic oscillator representation to handle non-Gaussian disorder in path integrals is a significant technical innovation applicable to other disordered systems.

In conclusion, this work successfully solves the LSYK model, characterizing its phase diagram, chaotic properties, and thermodynamics, and establishing its role as a solvable proxy for sparse quantum systems with rich holographic implications.