---
title: Measurement-Induced Criticality
url: https://www.emergentmind.com/topics/measurement-induced-criticality
type: topic
---

# Measurement-Induced Criticality

Measurement-induced criticality refers to universal critical phenomena that emerge when the entanglement structure of monitored quantum systems undergoes a sharp transition as the rate of local measurements is varied. At the heart of this phenomenon is the competition between unitary dynamics, which generate many-body entanglement, and local measurements, which tend to suppress entanglement and induce collapse. As the measurement rate passes through a critical value, the system exhibits a transition from a phase supporting volume-law entanglement to one obeying area-law scaling or even sub-volume entanglement, accompanied by emergent scale invariance and new universality classes distinct from conventional equilibrium criticality or classical percolation.

## 1. Fundamental Models and Mapping to Statistical Mechanics

The canonical models for measurement-induced criticality are hybrid quantum circuits consisting of spatially extended arrays (chains or lattices) of qubits (or higher-dimensional spins), evolving in discrete time via alternating layers of random unitary gates and stochastic projective measurements applied independently to each site with probability $p$ per update. The quantum dynamics of these circuits are described not by the ensemble-averaged density matrix—which always becomes maximally mixed—but by averages over pure-state trajectories conditioned on specific projective measurement outcomes.

The analytic understanding of measurement-induced entanglement transitions is established by mapping the averaged Rényi entropies of subregions to a classical statistical mechanics problem in $d+1$ dimensions (with $d$ the spatial dimension and the extra direction being the temporal or circuit depth) [1908.08051]. The replica-trick and permutation group structure underlying the average leads to $S_n$-spin models with effective temperature parameterized by $p$. In particular limits (large onsite Hilbert-space $d\to\infty$, Hartley entropy $n\to0$, or large bond dimension in random tensor networks), the model maps exactly onto classical percolation, but at finite $d$ and $n\ge1$ the universality class deviates in subtle but crucial ways from pure percolation.

## 2. Entanglement Transitions and Critical Scaling Laws

Within these hybrid circuits, the monitored dynamics support two competing phases:

- **Volume-law phase ($p<p_c$):** The unitary evolution dominates, leading to extensive entanglement entropy $S_A \sim |A|$ for a subregion $A$.
- **Area-law phase ($p>p_c$):** Frequent projective measurements localize information, so $S_A$ obeys area-law scaling or even saturates to $O(1)$. This is sometimes termed the quantum Zeno phase.

At the critical measurement rate $p=p_c$, the system exhibits universal scaling behavior ("measurement-induced criticality"):

- In $(1+1)$D, at $p=p_c$ the entanglement entropy of a region of length $\ell$ scales logarithmically: $S(\ell) \sim b_n \log\ell$, with a universal coefficient $b_n$; for example, in the percolation limit $b_n = 1/3$ for unitary circuits [1908.08051].
- In $(2+1)$D, at $p_c$ the leading area law receives a *multiplicative* logarithmic correction: $S(A) \sim a|\partial A|\log|\partial A|$ [2007.02970].
- The mutual information of distant subregions shows universal power-law decay at criticality, with scaling exponents inherited from the corresponding statistical mechanics mapping [1908.08051].
- The transition is characterized by a diverging correlation length $\xi \sim |p-p_c|^{-\nu}$, with precise exponents extracted from finite-size scaling analyses.

Numerically, for $(1+1)$D hybrid Clifford circuits, the critical point occurs at $p_c \simeq 0.16$–$0.17$, with correlation length exponent $\nu \approx 1.3$, and dynamical exponent $z=1$ [1910.00020, 2101.06245, 2012.03857]. In $(2+1)$D, for Clifford circuits with plaquette random unitaries, $p_c$ ranges from $0.54$ to $0.84$ depending on measurement protocol, and $\nu \approx 0.67$–$0.68$—distinct from the $\nu_{3D\text{-perc}}\approx 0.88$ of 3D percolation [2007.02970].

## 3. Field-Theoretic Structure and Percolation Analogy

The correspondence to percolation and statistical field theory is central for both analytic and numerical results:

- In the $d\to\infty$ or large bond dimension limit, the stat-mech model maps to bond percolation, with a transition at $p_c=1/2$ and critical exponents such as $\nu_{perc}=4/3$ in two dimensions. The entanglement entropy at $p=p_c$ is set by boundary-condition-changing (BCC) operators in $c=0$ conformal field theory, which yields universal logarithmic scaling of the entropy and mutual information [1908.08051].
- For finite onsite Hilbert space $d$, the percolation fixed point is perturbed by relevant operators ("two-hull" field with dimension $\Delta_{2\text{h}}=5/4$), resulting in a new universality class with different critical exponents, but still with emergent conformal or scale-invariant behavior [1908.08051, 2012.03857, 2007.02970].

Explicitly, the entanglement entropy for an interval of length $L_A$ at criticality in the percolation CFT is
\[
\overline{S_{n,A}}|_{p_c} = a_n L_A + b_n\log L_A + ...
\]
with $b_n = \frac{1}{3}$ for $n\geq1$ in the infinite-$d$ limit. For finite $d$, this coefficient is modified, with $b_n\approx 0.28$–$0.35$ observed in Clifford and Haar-random circuits [1908.08051].

Mutual information between two intervals $A, B$ a distance $r\gg1$ apart decays as $\sim r^{-2/3}$ at criticality in the percolation regime, an exact result derived from the conformal field theory mapping [1908.08051].

## 4. Universality Class and Deviations from Percolation

Measurement-induced entanglement transitions generically do not fall strictly within the classical percolation universality class at finite Hilbert-space dimension. For example, simulations of $(1+1)$D Clifford and Haar circuits, as well as $(2+1)$D Clifford circuits, yield critical exponents that are close to, but measurably distinct from, their percolation counterparts [2007.02970, 2012.03857]. Notably, for $(2+1)$D Clifford circuits:
- $\nu=0.67(1)$ (rank-1 measurements), $\nu=0.68(1)$ (rank-2), both $\sim20\%$ smaller than percolation.
- At criticality, entanglement scaling is $S(A) \sim c L\log L$, multiplicative rather than additive logarithmic violation [2007.02970].
- This establishes a novel universality class for non-equilibrium criticality in monitored quantum circuits, with bulk exponents close to percolation but distinct surface or boundary scaling, and different exponents for entanglement cluster observables [2012.03857].

These differences are understood as arising from the breaking of the full $S_{Q!}$ symmetry of the Potts/percolation point down to $S_Q\times S_Q$ by finite-$d$ corrections, with the resultant RG flow leading to a new $c=0$ field theory fixed point [1908.08051].

## 5. Extensions, Observable Probes, and Experimental Signatures

Measurement-induced criticality is robust across a range of models, statistics, and measurement protocols:

- Entanglement negativity, mutual information, and bipartite fluctuations can serve as sharp diagnostics of the entanglement transition and criticality, with negativity exponents ("mutual negativity") displaying possible super-universality across hybrid circuit families [2012.00031].
- Numerical and analytical studies confirm that subregion entanglement and mutual information exhibit finite-size scaling collapse near the critical point. The critical point can be located and its critical exponents extracted via scaling collapses of these order parameters [1910.00020, 2101.06245].
- Measurement-induced criticality persists—and can be sharply diagnosed—when generalized to higher spatial dimension, long-range and power-law interactions, in the presence of symmetry constraints, and in various measurement protocols (including continuous measurements and post-selected measurements) [2007.02970, 2107.05669, 2110.14403, 2603.15744].
- In random tensor networks, an analogous transition occurs, governed by a closely related $S_Q$-spin model; the resulting universality class depends on the replica limit ($Q\to0$ for RTNs, $Q\to1$ for monitored circuits), resulting in closely related but distinct critical exponents and amplitude prefactors [1908.08051].
- Experimental schemes using charge or spin fluctuations, efficient quantum tomography, and continuous monitoring protocols have been developed to directly probe measurement-induced criticality in platforms such as quantum gas microscopes and NISQ devices [2302.14044, 2308.01653].

## 6. Impact of Disorder, Non-Unitary Monitoring, and Post-Selection

The stability and universality of measurement-induced criticality are sensitive to disorder and measurement protocols:

- In $1$D, quenched spatial disorder in measurement rates drives the transition to an infinite-randomness critical point with activated dynamical scaling (logarithmic time vs. spatial size), steady-state entanglement scaling as $S(\ell)\sim\sqrt{\ell}$, and correlation length exponent $\nu=2$ [2205.14002]. This aligns with the Harris-Chayes bound and is generalizable to the entire family of disordered measurement protocols [2308.03844].
- Quasiperiodic modulation of measurement rates provides a deterministic mechanism to induce "infinite quasiperiodic" criticality, with scaling exponents continuously tunable via the wandering exponent $\beta$ of the QP sequence. The resulting critical points interpolate between conformal MIPT ($\nu\approx1.28$, $z=1$) and infinite-randomness universality, with scaling $S\sim L^\beta$ and $\xi\sim|p-p_c|^{-1/(1-\beta)}$ [2308.03844].
- Post-selected measurement protocols (forced outcomes) drive the system into a new universality class, with enhanced correlation length exponent $\nu\approx2.1$ and negative effective central charge $c_{\mathrm{eff}}\approx -0.4$, matching the transition in random tensor networks and confirming the sensitivity of universality class to the conditional statistics of the measurement outcomes [2603.15744].

## 7. Broader Context: Classical-Quantum Correspondence and Open Questions

Measurement-induced criticality provides a paradigm for non-unitary, non-equilibrium phase transitions that are both analytically tractable (via stat-mech mappings) and accessible to large-scale numerical simulation (e.g. stabilizer circuits via the Gottesman–Knill theorem). The foundational correspondence to classical percolation and $c=0$ field theories provides a powerful framework for understanding and classifying new universality classes emerging in monitored quantum dynamics [1908.08051, 2012.03857].

Key open questions include:
- The precise nature of the field theory describing the finite-$d$ (Clifford and qubit) universality class.
- The origin of the surprising proximity to percolation fixed points in certain models, and the distinction between bulk and boundary critical exponents.
- The detailed nature of entanglement cluster statistics and the geometric structure of entanglement at criticality in higher dimensions [2012.03857].

Measurement-induced criticality stands as a fundamental feature of monitored quantum many-body systems, enabling the systematic exploration of novel non-equilibrium universality classes and guiding the design of experiments and protocols in near-term quantum devices [2007.02970, 2308.01653, 2302.14044].

Source: https://www.emergentmind.com/topics/measurement-induced-criticality