⚛️ quantum physics

Spiral renormalization group flow and universal entanglement spectrum of the non-Hermitian 5-state Potts model

This paper demonstrates that tensor network algorithms can successfully simulate the non-Hermitian 5-state Potts model to observe the predicted spiral renormalization group flow and reconstruct the universal entanglement spectrum, thereby validating their utility for capturing the approximate conformal invariance of weakly first-order phase transitions described by complex conformal field theory.

Original authors: Vic Vander Linden, Boris De Vos, Kevin Vervoort, Frank Verstraete, Atsushi Ueda

Published 2026-04-01

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Vic Vander Linden, Boris De Vos, Kevin Vervoort, Frank Verstraete, Atsushi Ueda

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 understand how a complex system, like a crowd of people or a magnetic material, behaves when it's on the very edge of changing its state. Usually, scientists use a map called the Renormalization Group (RG) to predict these changes. Think of this map as a GPS for physics: it tells you how the "rules" of the system change as you zoom out from looking at individual atoms to looking at the whole material.

Normally, this GPS leads you to a "fixed point"—a destination where the rules stop changing and the system becomes perfectly balanced (like a calm lake). This is what happens in smooth, continuous phase transitions.

But what if the system is about to snap suddenly, like a brittle stick breaking? This is a first-order phase transition. For a long time, physicists thought these were just messy jumps with no underlying order. However, this paper suggests a fascinating secret: even in these sudden jumps, there is a hidden, ghostly order.

Here is a simple breakdown of what the authors discovered:

1. The "Ghost" Destination (Complex Fixed Points)

Imagine you are hiking toward a mountain peak (the fixed point). In normal physics, the peak is right in front of you. But in this specific model (the 5-state Potts model), the peak has vanished into a parallel dimension. It exists in a "complex" world—a mathematical space involving imaginary numbers.

Because the peak is in this ghostly dimension, you can't actually reach it. Instead, as you hike, you get stuck in a loop. You walk in circles around the invisible peak, getting closer and closer but never arriving. This is called "walking." It looks like the system is about to change smoothly, but it's actually just circling a ghost.

2. The Spiral Path

The authors found that if you plot the path of this "walking" on a map, it doesn't just go in a circle; it forms a spiral.

The Analogy: Imagine a squirrel running around a tree, but with every step, it gets slightly closer to the trunk while also spinning. That spiral path is the "RG flow" the paper describes.
The Discovery: By using powerful computer simulations (Tensor Networks), they were able to trace this spiral path for the first time on a large scale, confirming that the system is indeed circling these invisible, complex fixed points.

3. The "Non-Hermitian" Twist

To study this ghostly peak, the scientists had to tweak the rules of the game. They added a special ingredient to their mathematical model that made it non-Hermitian.

The Analogy: Think of a normal mirror (Hermitian) where the reflection is perfect and real. A non-Hermitian mirror is like a funhouse mirror that distorts reality, creating reflections that are slightly "off" or imaginary.
The Problem: Usually, computer algorithms break when you use these funhouse mirrors because the math gets messy. The "variational principle" (a standard rule for finding the best answer) stops working.
The Solution: The authors showed that because the "ghost" is so close to the real world (the system is only weakly first-order), the distortion is small. They proved that standard computer tools (Tensor Networks) could still handle the funhouse mirror and find the correct answer, provided they looked at large enough systems.

4. Listening to the "Entanglement Spectrum"

In quantum physics, particles can be "entangled," meaning they are linked across space. When you cut a system in half, the way the two halves are linked contains a secret code called the entanglement spectrum.

The Analogy: Imagine the system is a song. The "entanglement spectrum" is the sheet music that tells you exactly what notes are being played.
The Result: Even though the system is in this weird, complex state, the sheet music they found matched the theoretical predictions for a "boundary Conformal Field Theory." This means the hidden ghostly order leaves a clear fingerprint on the system's quantum connections.

Why Does This Matter?

This paper is a big deal because it proves that Tensor Networks (a type of super-smart computer algorithm) are the right tool to solve these tricky, non-Hermitian problems.

Before: We thought these "weakly first-order" transitions were just messy jumps.
Now: We know they are actually governed by a beautiful, spiraling dance around invisible, complex fixed points.
The Future: This gives us a new way to understand materials that change state abruptly. It suggests that even when things look chaotic or sudden, there is a deep, hidden mathematical harmony (conformal invariance) driving them, even if that harmony lives in a "complex" world we can't directly touch.

In a nutshell: The authors used advanced computer simulations to prove that a specific magnetic model doesn't just "jump" when it changes state; it actually spirals around a hidden, ghostly destination in a complex mathematical world, leaving a clear trace of this behavior in the quantum connections between its particles.

1. Problem Statement

The paper addresses the theoretical challenge of characterizing weakly first-order phase transitions, specifically in the quantum 5-state Potts model ( $Q=5$ ).

Context: While the $Q \leq 4$ Potts models undergo continuous phase transitions described by real Conformal Field Theories (CFT), the $Q > 4$ case is known to be first-order. However, the $Q=5$ case exhibits "walking" behavior (slow Renormalization Group flow) over intermediate scales, suggesting the presence of nearby complex fixed points.
Theoretical Framework: These fixed points are described by a Complex Conformal Field Theory (CCFT), which exists in a complexified coupling space. The model requires a non-Hermitian deformation to access these fixed points on the lattice.
Challenge: Standard numerical methods like Exact Diagonalization (ED) are limited to small system sizes ( $L \sim 12$ ), which is insufficient to observe the logarithmic spiral RG flow and finite-size scaling required to confirm the CCFT nature. Furthermore, standard tensor network algorithms (like DMRG) are designed for Hermitian systems; applying them to non-Hermitian Hamiltonians where the variational principle breaks down is non-trivial.

2. Methodology

The authors employ Tensor Network (TN) simulations, specifically Matrix Product States (MPS) and the Density Matrix Renormalization Group (DMRG) algorithm, to simulate the deformed non-Hermitian Hamiltonian.

Hamiltonian: The system is defined by the 5-state Potts Hamiltonian deformed by a non-Hermitian term $\lambda \hat{H}_1$ :
$\hat{H} = \sum_i \sum_{k=1}^{Q-1} [J(\hat{Z}^\dagger_i \hat{Z}_{i+1})^k + h \hat{X}^k_i] + \lambda \hat{H}_1$
where $\hat{H}_1$ is a specific nearest-neighbor extension term respecting $Z_5$ symmetry and Kramers-Wannier duality.
Algorithmic Adaptation:
- The authors define the ground state as the eigenstate with the smallest real part of the eigenvalue.
- They utilize standard DMRG (without explicit left/right eigenvector adaptation) based on the hypothesis that the non-Hermitian deformation is sufficiently small (due to the "walking" nature of the theory) that the standard MPS ansatz remains robust.
- They enforce $Z_5$ symmetry to improve computational efficiency and identify operator charges.
- Simulations were performed on system sizes up to $L=64$ (for entanglement spectra) and $L=24$ (for precise RG flow), significantly exceeding the limits of ED.
Analysis Techniques:
- Finite-Size Scaling (FSS): Used to extract the critical parameter $\lambda_c$ and conformal data (central charge $c$ , scaling dimensions $\Delta_n$ ) by fitting energy gaps to CFT predictions on a cylinder.
- Cost Function Minimization: A cost function $J(v, g_{\epsilon'})$ was constructed to minimize the deviation between theoretical and estimated scaling dimensions to locate the fixed point.
- Entanglement Spectrum: The authors constructed a non-Hermitian reduced density matrix $\rho_{RL} = |\psi_L\rangle\langle\psi_R| / \langle\psi_R|\psi_L\rangle$ to extract the entanglement Hamiltonian $K_C$ and its spectrum.

3. Key Contributions

Validation of TN for Non-Hermitian CCFT: The paper demonstrates that standard DMRG/MPS algorithms, despite being designed for Hermitian systems, can accurately capture the physics of weakly non-Hermitian theories. The "walking" behavior ensures the non-Hermiticity is a small perturbation, preserving the ordering of singular values and entanglement structure.
High-Precision Critical Point: By utilizing larger system sizes and subleading finite-size scaling corrections, the authors refined the critical complex coupling parameter to:
$\lambda_c = 0.0788 + 0.0603i$
This represents a significant improvement in precision over previous estimates ( $0.079 + 0.060i$ ).
Observation of Spiral RG Flow: The study successfully mapped the Renormalization Group flow of the running coupling $g_{\epsilon'}$ in the complex plane, visually confirming the theoretically predicted logarithmic spiral trajectory around the complex fixed points.
Reconstruction of Boundary CCFT: The authors reconstructed the full boundary CCFT spectrum via the entanglement Hamiltonian, identifying the correct operator content and degeneracies corresponding to free-free boundary conditions.

4. Key Results

Critical Parameters: The refined critical point yields conformal data (central charge and scaling dimensions) that match theoretical predictions for the $Q=5$ $Q = 5$ CCFT with high accuracy.
- Central Charge: $c \approx 1.1375 - 0.0211i$ .
- Scaling dimensions for operators like $\sigma, \epsilon, \sigma'$ were extracted and found to be consistent with theory, though the imaginary parts showed slightly slower convergence due to MPS truncation artifacts.
RG Flow Topology: The flow of the coupling parameter $g_{\epsilon'}$ exhibits a clear spiral trajectory in the complex theory space, converging toward the fixed points $C$ and $\bar{C}$ . This confirms the "walking" regime is governed by complex fixed points.
Entanglement Spectrum:
- The entanglement spectrum of the lattice model (with Periodic Boundary Conditions) flows to the free-free Conformal Boundary Condition (CBC) spectrum.
- The spectrum displays the correct degeneracies (1, 4, and 11) for the primary operators $I, \phi_{3,1}, \phi_{5,1}$ , confirming the emergence of conformal invariance on the lattice.
- The imaginary part of the spectrum converges more slowly, attributed to the non-positive-definite nature of the non-Hermitian density matrix and MPS truncation errors in the imaginary sector.

5. Significance

Methodological Breakthrough: This work establishes tensor networks as a viable and powerful tool for studying non-Hermitian critical phenomena and complex CFTs, overcoming the limitations of exact diagonalization.
Physical Insight: It provides strong numerical evidence that weakly first-order transitions in the 5-state Potts model are indeed governed by complex fixed points, validating the "walking" scenario and the existence of CCFTs in lattice models.
Universal Scaling: The successful extraction of the entanglement spectrum and RG flow demonstrates that conformal invariance emerges even in non-Hermitian systems, bridging the gap between lattice simulations and complex field theory.
Future Directions: The authors suggest that these methods could be applied to other open questions, such as the nature of the Néel-Valence Bond Solid transition and the search for complex fixed points in other deconfined quantum critical points. They also highlight the need for infinite-size tensor network algorithms to further reduce finite-size effects.

In summary, the paper successfully combines advanced tensor network techniques with complex field theory to numerically verify the existence and properties of complex fixed points in the 5-state Potts model, providing a new paradigm for simulating non-Hermitian quantum criticality.