⚛️ quantum physics

Elusive phase transition in the replica limit of monitored systems

This paper demonstrates that in an exactly solvable model of monitored quantum dynamics, non-perturbative logarithmic corrections arising in the physically relevant replica limit ( $n\to1$ ) eliminate the purifying phase observed at finite replica numbers, thereby ensuring that the purification time remains exponentially long in system size regardless of the measurement rate.

Original authors: Guido Giachetti, Andrea De Luca

Published 2026-07-07

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: Guido Giachetti, Andrea De Luca

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 a giant room filled with $N$ tiny, spinning tops (quantum particles). These tops are constantly bumping into each other in a chaotic, noisy way, trying to mix up their states. At the same time, a "watcher" is constantly peeking at them, taking weak, fuzzy snapshots of their positions from random angles.

In the world of quantum physics, there's a famous debate about what happens when you have this mix of chaos and watching:

The Chaos: If you just let them spin, they get so tangled up that they become a giant, messy cloud of information (high entanglement).
The Watching: If you watch them too closely, the act of looking forces them to pick a definite state, cleaning up the mess (purification).

Scientists have long suspected there is a "tipping point" (a phase transition). Below a certain level of watching, the tops stay messy. Above it, they snap into a clean, ordered state.

The Problem with the "Math Trick"
To study this, physicists use a clever mathematical shortcut called the Replica Trick. Imagine you don't just have one room of tops, but you make $n$ identical copies (replicas) of the whole room. You do the math on these copies, and then, at the very end, you try to shrink the number of copies down to exactly one ( $n \to 1$ ) to see what happens in the real world.

For a long time, calculations using this trick suggested that if you watch hard enough, the system would snap into a clean state. It looked like a clear switch: messy on one side, clean on the other.

The Surprise Discovery
This paper by Guido Giachetti and Andrea De Luca takes a specific, solvable model of these spinning tops and runs the math all the way to the finish line. They found something shocking: The switch doesn't actually exist.

Here is the analogy they use to explain why:
Imagine you are trying to walk from a "Messy Room" to a "Clean Room."

When you look at the math with 2 or 3 copies ( $n=2, 3$ ), there is a clear door between the rooms. If you push hard enough (high measurement rate), you can walk through the door and get to the Clean Room quickly.
However, the real world is the limit of 1 copy ( $n=1$ ). When the authors zoomed in on this specific limit, they found that the "door" wasn't just closed; it was replaced by a foggy, logarithmic wall.

This wall is made of "non-perturbative logarithmic corrections." In plain English, this means that as you try to get closer to the real world (1 copy), the math changes in a way that creates an invisible, sticky barrier. No matter how hard you watch (how high the measurement rate is), this barrier prevents the system from ever fully cleaning itself up quickly.

The Result: An Endless Wait
Because of this invisible wall, the time it takes for the system to go from a messy, mixed state to a clean, pure state is exponentially long.

Think of it like this:

The Old View: If you watch hard enough, the mess clears up in a few seconds.
The New View: Even if you watch with a super-microscope, the mess will take so long to clear up that it might as well be forever. The time required grows so fast with the size of the system that for a large system, you would have to wait longer than the age of the universe to see it clean up.

Why This Matters
The authors show that the "phase transition" (the switch from messy to clean) is an illusion created by the mathematical shortcut when you stop at 2 or 3 copies. The "real" physics (the $n \to 1$ limit) is much more stubborn. The system stays in a "volume-law" phase (messy and entangled) no matter what, because the mathematical corrections that appear only at the very end of the calculation destroy the "clean" phase entirely.

In short: In this specific model, you can never win the game of purification. The system is too stubborn to be cleaned up by watching, no matter how much you watch.

Technical Summary: Elusive Phase Transition in the Replica Limit of Monitored Systems

Problem Statement
The paper investigates the out-of-equilibrium dynamics of monitored quantum many-body systems, specifically focusing on the existence of measurement-induced phase transitions (MIPTs). While numerical studies and solvable models suggest a transition between a "volume-law" entangled phase (weak measurement) and an "area-law" phase (strong measurement), a rigorous analytical understanding remains elusive. A central difficulty is the nonlinearity introduced by the Born rule, which weights measurement trajectories by their probability. Standard approaches utilize the "replica trick" to linearize this problem by studying $n$ copies of the system, followed by an analytic continuation to the physical limit $n \to 1$ . However, subtleties in this $n \to 1$ limit—such as non-perturbative effects—are often inaccessible at finite integer $n$ , leading to potential discrepancies between finite- $n$ predictions and physical reality.

Methodology
The authors introduce an exactly solvable model of monitored dynamics consisting of $N$ spin-1/2 particles with all-to-all isotropic noisy interactions, where each spin is continuously weakly measured along a random direction.

Replica Formalism: To handle the Born-rule weighting, the authors employ the replica trick. They consider $n$ replicas of the system and derive an effective action for the replicated density matrix.
Path Integral and Mean-Field Theory: Using a Keldysh spin-coherent path integral, they reduce the many-body problem to a single-site (0+1) dimensional theory via Hubbard-Stratonovich fields. In the large- $N$ limit, the path integral is dominated by saddle-point solutions.
Symmetry Analysis: The effective action possesses a $(S_n \times S_n) \rtimes Z_2$ symmetry. The authors analyze symmetry-breaking sectors labeled by permutations $P \in S_n$ , which are crucial for describing purification dynamics (specifically the transition between the identity sector and the transposition sector).
Analytic Continuation: A key methodological step involves solving the saddle-point equations for arbitrary integer $n$ and then performing an exact analytic continuation to the physical limit $n \to 1$ . This is achieved by asymptotically expanding the largest eigenvalue of the transfer matrix and resumming the series, revealing non-perturbative logarithmic corrections.

Key Contributions and Results

Discrepancy between Finite $n$ and $n \to 1$ : The study finds that the nature of the phase transition depends critically on the number of replicas $n$ $n$ . For finite integer $n \ge 2$ $n \geq 2$ , the system exhibits a phase transition:
- For $n=2$ , there is a second-order transition at a critical measurement rate $\gamma_c = 1/4$ .
- For $n \ge 3$ , the transition is discontinuous.
- In these finite- $n$ cases, a "purifying phase" (where the system purifies in polynomial time) exists for sufficiently large $\gamma$ .
Absence of Transition in the Physical Limit: In the physically relevant limit $n \to 1$ , the free energy displays non-perturbative logarithmic corrections ( $j(x) \sim x^2 \ln x$ ). These corrections stabilize the volume-law phase for any finite measurement rate $\gamma$ . Consequently, the "purifying phase" present at finite $n$ disappears entirely.
Exponential Purification Time: As a direct consequence of the absence of a phase transition in the $n \to 1$ limit, the purification time for an initially mixed state is always exponentially long in the system size ( $T_{pur} \sim e^{cN}$ ), regardless of how large the measurement rate is. The system never enters a true area-law phase where purification is rapid.
Validation via Local Observables: The authors validate their mean-field theory by deriving the full statistics of single-site observables. They demonstrate that the boundary distribution of local spin components (relevant for observables) differs from the bulk stationary law due to transient effects, yet the analytical predictions match exact finite- $N$ numerical dynamics.

Significance and Claims
The paper claims that the "elusive" nature of the phase transition in monitored systems arises from the non-commutativity of the limits $x \to 0$ (order parameter) and $n \to 1$ (replica limit).

Limitation of Standard Approaches: The results suggest that standard Landau-Ginzburg field theories and perturbative renormalization group approaches, which rely on regular expansions around unbroken saddles at integer $n$ , may fail to capture the true physics of monitored systems in the physical limit.
Non-Perturbative Mechanism: The disappearance of the transition is attributed to non-perturbative logarithmic corrections emerging specifically at $n \to 1$ , potentially linked to the degeneracy of anomalous dimensions of operators in this limit.
Model Specificity vs. Generality: While the authors present an exactly solvable model, they pose the question of whether this logarithmic non-analyticity is a general feature of monitored systems or a peculiarity of the all-to-all connectivity. They suggest that if general, it would fundamentally alter the understanding of MIPTs and the classical simulatability of such dynamics.

The work concludes that for the specific model studied, the measurement-induced phase transition is an artifact of the finite- $n$ approximation, and the physical system remains in a "replica-broken" phase with exponentially long purification times for all measurement strengths.

Technical Summary: Elusive Phase Transition in the Replica Limit of Monitored Systems

More like this