Rare Events Govern Defect Formation under Weak Symmetry Breaking

arXiv:2606.27835 · cond-mat.stat-mech · Submitted 2026-06-26 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Condensed Matter. The day in condensed matter: materials, phases and the measurements under them.

Kai: We're looking at the paper: "Rare Events Govern Defect Formation under Weak Symmetry Breaking".

Mira: Here's the gist of the paper. Rare events govern defect formation under weak symmetry breaking by showing that defect formation is controlled by rare fluctuations that drive local regions into the disfavored symmetry-broken state <ref:2606.27835#pg2> The resulting defect density is obtained by multiplying this probability by the usual density of independent KZ domains, n(d) defects ∼ η τQ d 4 exp

−Ad h squared / (θ τQ η d+2/4): This theoretical framework is verified through direct simulations of stochastic Ginzburg–Landau models in one and two spatial dimensions The Kibble-Zurek Mechanism and its Limitations Crossing a continuous phase transition out of equilibrium typically generates topological defects whose density obeys a universal power-law scaling predicted by the Kibble–Zurek mechanism <ref:2606.27835#pg2> The number of defects scales as nKZ ∝ τ- dν 1+zν Q, (1) where d is the spatial dimensionality of the system, and ν and z are the equilibrium correlation-length and dynamical critical exponents that define the universality class of the transition <ref:2606.27835#pg2> This prediction has been verified in a broad range of experimental systems, including colloids, superfluid helium, liquid crystals, and ultracold atomic gases <ref:2606.27835#pg2> More recently, considerable attention has focused on the fate of the Kibble–Zurek mechanism in systems that weakly depart from the paradigm of continuous phase transitions <ref:2606.27835#pg2> In both cases, systematic deviations from the standard KZ scaling in Eq.

Kai: (1) have been observed <ref:2606.27835#pg2>.

Mira: The Role of Weak Symmetry Breaking When a weak symmetry-breaking field h is introduced, the symmetry-breaking choice of each KZ domain is no longer completely random <ref:2606.27835#pg2> Instead, the external field biases the local selection of the order parameter, making the process analogous to a biased coin toss that favors the symmetry-broken state <ref:2606.27835#pg2>.

Kai: The probability of forming a defect is no longer determined by a quench-rate-independent random mismatch between neighboring domains, but by the probability that a domain is driven into the disfavored state against the bias imposed by the external field <ref:2606.27835#pg2> This physical origin is illustrated in Fig.

Title and authors: Kai: So we're looking at this paper, "Rare Events Govern Defect Formation under Weak Symmetry Breaking." It looks like they’re taking something we know about how defects form when a system crosses a phase transition out of equilibrium, which is usually described by the Kibble-Zurek mechanism.

Mira: Exactly. The main point here is that when you add a weak symmetry-breaking field, it doesn't just change things slightly; it shifts the entire picture from something predictable based on random mismatches to something controlled by rare fluctuations.

Lev: From an engineering standpoint, that means we can't just rely on the standard power-law scaling anymore if we want to predict defect densities accurately in systems where perfect symmetry isn't achievable.

Kai: The paper explains that defect formation is now controlled by those rare events that drive local regions into the disfavored state against some external bias, which is what they call a biased coin toss analogy in Figure one <ref:2606.27835#pg1>.

Mira: It’s not about whether a domain randomly picks the wrong state anymore; it's about the probability of that domain being driven into the wrong state because of that external field.

Lev: So, instead of just looking at the quench rate and dimensionality, we have to consider how much noise is needed for a fluctuation to overcome that bias, which they call Pwrong.

Kai: They derive a specific scaling for this rare fluctuation action cost, and then multiply it by the usual density of domains to get the final defect density prediction.

Mira: And the result they get is an exponential correction term on top of that standard Kibble-Zurek power law scaling, which is pretty significant because it shows how weak symmetry breaking impacts things at a deeper level.

Lev: If we want to run this on real hardware, the action cost they derive, Sh, scales in a way that depends on the dimensionality d and some other parameters like theta and tau Q.

Kai: The paper verifies this theory numerically using stochastic Ginzburg-Landau models in both one and two spatial dimensions, showing that the exponential correction is consistent across those different settings.

Mira: They also show a universal collapse of a parameter, lambda, onto a single scaling form predicted by Eq. five for both the one and two dimensional cases, which is really nice because it suggests the underlying physics isn't changing much between these symmetry groups <ref:2606.27835#pg3>.

Lev: For us building error correction codes or simulating phase transitions on real hardware, this means that if we have a weak bias field h, we can use this formula to estimate how many defects we expect to see, and it gives us an exponential suppression factor.

Kai: The main implication for the experimental community is that they've provided a closed-form expression for the defect density in arbitrary dimensions under these conditions.

Mira: It moves us away from just guessing what happens when symmetry breaks weakly, giving us a concrete mathematical tool to calculate these rare event corrections.

Title and authors: Lev: It tells us exactly how much of the defect formation is due to those rare fluctuations that manage to overcome the bias imposed by h and theta.

Kai: So, summarizing this paper, it shows that weak symmetry breaking transforms defect formation from a simple Kibble-Zurek process into a rare event problem governed by an exponential suppression factor.

Mira: That exponential term is controlled by the scaling of the action cost Sh, which they show has a specific dependence on the quench time tau Q and dimensionality.

Lev: And for practical application, it means that when we consider materials or systems with slight symmetry violations, we need to account for this rare event physics to get accurate numbers.

Kai: So what does this mean for us moving forward? It suggests that these results are directly applicable to real-world phase transitions, like those in quantum superfluids where perfect symmetry is rarely realized.

Mira: And the paper points out that symmetry-violating biases have already been observed experimentally to suppress the Kibble-Zurek mechanism in quantum superfluids, which lends some credence to this theoretical work.

Lev: It's testable right now, because they’ve given us a clear scaling relation that we can plug parameters into and compare against what we measure in simulations or experiments.

Kai: I think the biggest change here is shifting our focus from just the average defect density predicted by Kibble-Zurek to understanding the probability of those rare, bias-driven events.

Mira: It gives us a way to calculate that probability directly through that action cost Sh, which is derived from large deviation theory and the Freidlin-Wentzell action functional.

Lev: That means we can use this framework to identify and predict metastable configurations in complex systems where defects are governed by overcoming energy barriers rather than just random spatial mismatches.

Kai: And the unification between the Z2 and U1 models under this framework is also important, showing that the core scaling exponent remains consistent even when you change the symmetry group.

Mira: That consistency across different symmetry groups, like Ising and U(one) models, suggests a more fundamental structure to how these topological defects behave in Ginzburg-Landau dynamics <ref:2606.27835#pg1>.

Lev: From an error correction perspective, it implies that the way noise interacts with the system under bias is universal enough to predict defect statistics regardless of the specific field configuration we're dealing with.

Kai: To wrap this up, this paper on "Rare Events Govern Defect Formation under Weak Symmetry Breaking" gives us a concrete tool to calculate those exponential corrections in defect densities when symmetry is weakly broken.

Mira: It’s a big step because it moves us beyond the standard Kibble-Zurek power law and shows how rare fluctuations dictate the outcome.

Lev: We need to keep an eye on how this theory extends beyond the mean-field approximation, because that's where we get closer to modeling real hardware with its full complexity.

Kai: That’s what it is—a tool for calculating those rare event corrections in defect densities under weak symmetry breaking.

The paper's summary: Kai: So, we're looking at how this paper summarizes its main finding—that weak symmetry breaking turns defect formation from something predictable, like the Kibble-Zurek mechanism, into a rare event problem.

Mira: Exactly. It basically says that when you introduce that external bias field, the system doesn't just get a slight nudge; it has to overcome an energy barrier by chance for a defect to form.

Lev: Which means we stop thinking about simple random mismatches between neighboring domains and start thinking about the probability of those rare fluctuations actually succeeding in driving one domain into the wrong state.

Kai: The key result they land on is this formula for the defect density, which shows an exponential correction on top of the standard power law scaling you already know from Kibble-Zurek.

Mira: That exponential term depends on how strong that external bias field h is, and also how much noise there is in the system, which they call theta.

Lev: From an engineering standpoint, that means we can use this scaling relation to calculate exactly how much the defect density will be suppressed by a specific physical parameter in our simulations or experiments.

Kai: The authors verified this mathematically using stochastic Ginzburg-Landau models in both one and two spatial dimensions, which gives us strong evidence that this isn't just some theoretical trick.

Mira: They also showed a nice feature called a universal collapse of the parameter lambda onto a single scaling form for both the two dimensional system and the one dimensional system.

Lev: That consistency across different dimensions is important because it suggests that the core physics—the way those rare fluctuations scale—is universal, even when you change the symmetry group, like going from Ising to U one models.

Kai: It also unifies how Z two and U one topological defects behave under Ginzburg-Landau dynamics, showing a single theoretical structure for both.

Mira: So what this really means for us is that we have a concrete tool now to calculate the exponential corrections when symmetry is weakly broken in any dimension, which is much more realistic than just relying on the standard Kibble-Zurek scaling.

Lev: It means if we're designing materials or simulating quantum systems where perfect symmetry isn't achieved, this framework gives us a way to predict defect suppression as a function of the bias field and the noise level.

Kai: And it’s testable now because they show this agreement with numerical simulations in both one and two dimensions.

Mira: The next thing we have to look at is how this framework applies when you move beyond just the mean-field approximation, because that’s where we really need to go for real hardware.

The paper's improvements: Kai: So, we're looking at what they suggest improving in this paper—how they want to push this research forward beyond what’s currently there about weak symmetry breaking and defect formation.

Mira: They point out that the current work is mostly in the mean-field approximation, so the next big step would be extending this framework to include those beyond-mean-field corrections.

Lev: That makes sense because for real hardware, you can't ignore those higher order terms; they land right where your error correction overhead really lands.

Kai: They suggest applying the derived scaling relation, that defect density formula, directly to realistic phase transitions where perfect symmetry is rarely realized in practice.

Mira: It’s about making this concrete tool immediately testable, so instead of just theory on a page, we can plug in parameters and see if it matches what we measure in simulations or experiments.

Lev: I agree with that; having a predictive scaling relation means we can use it to design materials where defects are suppressed by specific physical parameters like the bias field h or the noise strength theta.

Kai: They also highlight that they need to focus on how this applies when we look at systems with different symmetry groups, like comparing Z two and U one models.

Mira: That unification is important because it shows that the core scaling exponent remains consistent across different symmetry groups under Ginzburg-Landau dynamics, which is a deeper structural finding than just getting the final number.

Lev: If we can generalize this to handle more complex systems, it opens up possibilities for AI models to predict topological features in a much wider variety of physical settings.

Kai: So, the main implication is that this isn't just an interesting theoretical exercise; it’s a practical roadmap for how we should approach defect prediction in real-world physical systems.

Mira: They’re essentially saying that this rare event approach is the right way to handle weak symmetry breaking, and it needs to be pushed further into more complex physics.

Lev: It really shows that understanding these rare fluctuations is a necessary step before we can build robust error correction codes for materials that have inherent symmetry violations.

Conclusion: Kai: So we've been talking about how this paper, "Rare Events Govern Defect Formation under Weak Symmetry Breaking," shows that weak symmetry breaking turns defect formation into a rare event problem controlled by exponential suppression.

Mira: Right. It’s not just the average density anymore; it's about those rare fluctuations overcoming the bias imposed by an external field h, and that probability scales with the action cost Sh.

Lev: From an error correction standpoint, that means we have a way to quantify exactly how much suppression we get based on parameters like theta and tau Q for our simulations.

Kai: The main implication is that this framework moves us away from just Kibble-Zurek scaling and gives us a concrete tool for predicting defect densities in realistic settings, especially where perfect symmetry isn't there.

Mira: They also showed that the Z two and U one models unify under this same large deviation origin, meaning the core scaling exponent is consistent regardless of the symmetry group you're looking at.

Lev: That consistency across different groups is a big deal because it suggests a more fundamental structure to how these topological defects behave in Ginzburg-Landau dynamics.

Kai: So, to wrap up, this paper gives us the formula for that exponential correction on top of the standard Kibble-Zurek scaling when symmetry is weakly broken.

Mira: It’s a big step because it shows how rare fluctuations dictate the outcome and provides a way to calculate that probability directly through that action cost Sh.

Lev: We need to keep an eye on this theory extending beyond just mean-field approximations, because that's where we get closer to modeling real hardware with its full complexity.

Kai: Exactly, because those higher order terms are what matter when you’re trying to build something that actually runs in the real world.

Mira: We’ll keep an eye on how this framework applies when you move into those more complex systems where symmetry isn't perfect.

Jiang Liu, Peng Yang, Matteo Baggioli

Wilczek Quantum Center · Shanghai Jiao Tong University · Shanghai Research Center for Quantum Sciences · The Chinese University of Hong Kong

cond-mat.stat-mech

Submitted: 2026-06-26

Updated: 2026-11-04

Comments: v3: extended analysis and references added

License: http://creativecommons.org/licenses/by/4.0/

Importance score: 91/100

The gist: Rare events govern defect formation under weak symmetry breaking by showing that defect formation is controlled by rare fluctuations that drive local regions into the disfavored symmetry-broken state…

Key concepts

Kibble-Zurek Mechanism (KZ)
This is the standard theory predicting how many topological defects form when a system crosses a phase transition out of equilibrium. It suggests the number scales with time and spatial dimension according to universal power laws, assuming random domain formation.
Weak Symmetry Breaking
Introducing an external field biases which symmetry-broken state a local region chooses during the phase transition. This bias means defects are no longer formed by simple random mismatches but require rare fluctuations to drive a domain into the disfavored state.
Freidlin–Wentzell Action Functional
This mathematical tool describes the probability of a system following a specific, rare path in the presence of noise. It quantifies the 'cost' or action required for a stochastic process to deviate from its typical deterministic trajectory.
Rare Event Problem
When symmetry breaking is weak, defect formation becomes governed by rare fluctuations that overcome an energy barrier imposed by the external bias. This means defects are not abundant but are instead determined by the probability of these specific, low-probability events occurring.

Terminology

Summary

Rare events govern defect formation under weak symmetry breaking by showing that defect formation is controlled by rare fluctuations that drive local regions into the disfavored symmetry-broken state <ref:2606. p. 0

Where the paper says it

However, the origin of this correction and a general theoretical framework to describe it have remained elusive. Here, using large-deviation theory, we show that defect formation under weak symmetry breaking is controlled by rare fluctuations that drive local regions into the disfavored symmetry-broken state.

27835#pg2> The resulting defect density is obtained by multiplying this probability by the usual density of independent KZ domains, n(d) defects ∼ η τQ d 4 exp [−Ad h squared / (θ τQ η d+2/4)] This theoretical framework is verified through direct simulations of stochastic Ginzburg–Landau models in one and two spatial dimensions

The Kibble-Zurek Mechanism and its Limitations

Crossing a continuous phase transition out of equilibrium typically generates topological defects whose density obeys a universal power-law scaling predicted by the Kibble–Zurek mechanism <ref:2606.27835#pg2> The number of defects scales as nKZ ∝ τ- dν 1+zν Q, (1) where d is the spatial dimensionality of the system, and ν and z are the equilibrium correlation-length and dynamical critical exponents that define the universality class of the transition <ref:2606. p. 0

Where the paper says it

As predicted by the Kibble–Zurek (KZ) mechanism [2–5], the density of defects produced during the transition follows a universal power-law dependence on the quench time τQ. For sufficiently slow quenches (see [6] for the fast quench regime) and defects of zero dimensionality, the number of defects scales as nKZ ∝ τ − dν 1+zν Q , (1) where d is the spatial dimensionality of the system, and ν and z are the equilibrium correlation-length and dynamical critical exponents that define the universality class of the transition.

27835#pg2> This prediction has been verified in a broad range of experimental systems, including colloids, superfluid helium, liquid crystals, and ultracold atomic gases <ref:2606. p. 0
Where the paper says it

For sufficiently slow quenches (see [6] for the fast quench regime) and defects of zero dimensionality, the number of defects scales as nKZ ∝ τ − dν 1+zν Q , (1) where d is the spatial dimensionality of the system, and ν and z are the equilibrium correlation-length and dynamical critical exponents that define the universality class of the transition. This prediction has been verified in a broad range of experimental systems, including colloids, superfluid helium, liquid crystals, and ultracold atomic gases [1].

27835#pg2> More recently, considerable attention has focused on the fate of the Kibble–Zurek mechanism in systems that weakly depart from the paradigm of continuous phase transitions <ref:2606. p. 0
Where the paper says it

The Kibble–Zurek mechanism is therefore widely regarded as one of the cornerstones of nonequilibrium physics. More recently, considerable attention has focused on the fate of the Kibble–Zurek mechanism in systems that weakly depart from the paradigm of continuous phase transitions, including weakly first-order transitions [7] and transitions subject to weak explicit symmetry breaking [8].

27835#pg2> In both cases, systematic deviations from the standard KZ scaling in Eq. p. 0
Where the paper says it

More recently, considerable attention has focused on the fate of the Kibble–Zurek mechanism in systems that weakly depart from the paradigm of continuous phase transitions, including weakly first-order transitions [7] and transitions subject to weak explicit symmetry breaking [8]. In both cases, systematic deviations from the standard KZ scaling in Eq.

(1) have been observed <ref:2606. p. 0
Where the paper says it

However, when a weak symmetry-breaking field h is introduced, the symmetry-breaking choice of each KZ domain is no longer completely random. Instead, the external field biases the local selection of the order parameter, making the process analogous to a biased coin toss that favors the symmetry-broken state correspondarXiv:2606.

27835#pg2>.

The Role of Weak Symmetry Breaking

When a weak symmetry-breaking field h is introduced, the symmetry-breaking choice of each KZ domain is no longer completely random <ref:2606.27835#pg2> Instead, the external field biases the local selection of the order parameter, making the process analogous to a biased coin toss that favors the symmetry-broken state <ref:2606. p. 0

Where the paper says it

However, when a weak symmetry-breaking field h is introduced, the symmetry-breaking choice of each KZ domain is no longer completely random. Instead, the external field biases the local selection of the order parameter, making the process analogous to a biased coin toss that favors the symmetry-broken state correspondarXiv:2606.

27835#pg2>. The probability of forming a defect is no longer determined by a quench-rate-independent random mismatch between neighboring domains, but by the probability that a domain is driven into the disfavored state against the bias imposed by the external field <ref:2606. p. 1
Where the paper says it

1(b), each KZ domain selects the favored state with probability P+ and the disfavored state with probability P−, with P+ > P−. As a result, the probability of forming a defect is no longer determined by a quench-rate-independent random mismatch between neighboring domains, but by the probability that a domain is driven into the disfavored state against the bias imposed by the external field.

27835#pg2> This physical origin is illustrated in Fig. p. 1
Where the paper says it

As a result, the probability of forming a defect is no longer determined by a quench-rate-independent random mismatch between neighboring domains, but by the probability that a domain is driven into the disfavored state against the bias imposed by the external field. The physical origin of this process is illustrated in Fig.

1(c) for a real scalar field ϕ, where an external field h explicitly breaks the original Z2 symmetry, tilting the free-energy landscape and lifting the degeneracy between the two symmetry-broken minima <ref:2606. p. 1
Where the paper says it

The physical origin of this process is illustrated in Fig. 1(c) for a real scalar field ϕ, where an external field h explicitly breaks the original Z2 symmetry, tilting the free-energy landscape and lifting the degeneracy between the two symmetry-broken minima, one stable and the other metastable.

27835#pg2> Consequently, the formation of a defect requires a rare fluctuation that overcomes the bias and drives a KZ domain into the disfavored minimum <ref:2606. p. 1
Where the paper says it

1(c) for a real scalar field ϕ, where an external field h explicitly breaks the original Z2 symmetry, tilting the free-energy landscape and lifting the degeneracy between the two symmetry-broken minima, one stable and the other metastable. Consequently, the formation of a defect requires a rare fluctuation that overcomes the bias and drives a KZ domain into the disfavored minimum (Fig.

27835#pg2>.

Theoretical Derivation via Large-Deviation Theory

The probability of such a fluctuation scales as Pwrong ≍ exp [−Sh] = exp − h squared / (4ηθ ΩKZ) <ref:2606.27835#pg2> The corresponding action cost is Sh ∼ h squared ηθ tˆ d = h squared θ / (τQ η ν(d+z) 1+zν) <ref:2606. p. 1

Where the paper says it

1(d), and therefore scales as Pwrong ≍ exp [−Sh] = exp − h 2 4ηθ ΩKZ . (3) The corresponding action cost is Sh ∼ h 2 ηθ tˆˆξ d = h 2 θ τQ η ν(d+z) 1+zν .

27835#pg2> Since defects can only arise from such rare domains, the defect density is obtained by multiplying this probability by the usual density of independent KZ domains, n(d) defects ∼ η τQ d 4 exp [−Ad h squared / (θ τQ η d+2/4)] <ref:2606. p. 1
Where the paper says it

(3) The corresponding action cost is Sh ∼ h 2 ηθ tˆˆξ d = h 2 θ τQ η ν(d+z) 1+zν . (4) Since defects can only arise from such rare domains, the defect density is obtained by multiplying this probability

27835#pg2> This is the main result of this work <ref:2606. p. 2
Where the paper says it

by the usual density of independent KZ domains, nKZ ∼ ˆξ −d , which finally yields n (d) defects ∼ η τQ dν 1+zν exp  −Ad h 2 θ τQ η ν(d+z) 1+zν   , (5) where Ad is an overall constant that depends on microscopic details, hence not universal. This is the main result of this work.

27835#pg2>.

Numerical Verification and Unification

Numerical verification shows a clear exponential correction to the standard KZ power-law scaling, consistent with the numerical findings of Ref. [8] (see Fig. p. 0

Where the paper says it

Since each domain chooses its symmetry-broken state independently and at random, defects form at the interfaces between neighboring domains that make incompatible choices. The resulting defect density is therefore proportional to the number of domains, ndefects ∝ Pmismatch Nblocks, where Pmismatch is the probability that two neighboring domains select different broken-symmetry states (see Fig.

2(b) for the 1D results) <ref:2606. p. 0
Where the paper says it

However, when a weak symmetry-breaking field h is introduced, the symmetry-breaking choice of each KZ domain is no longer completely random. Instead, the external field biases the local selection of the order parameter, making the process analogous to a biased coin toss that favors the symmetry-broken state correspondarXiv:2606.

27835#pg2> The action cost Sh governing the exponential suppression of defects exhibits the scaling Sh ∝ τ(d+2)/4Q predicted by Eq. p. 3
Where the paper says it

2(b) for the 1D results). We find that the action cost Sh governing the exponential suppression of defects exhibits the scaling Sh ∝ τ (d+2)/4 Q predicted by Eq.

(5) <ref:2606. p. 0
Where the paper says it

However, when a weak symmetry-breaking field h is introduced, the symmetry-breaking choice of each KZ domain is no longer completely random. Instead, the external field biases the local selection of the order parameter, making the process analogous to a biased coin toss that favors the symmetry-broken state correspondarXiv:2606.

p. 0
Where the paper says it

However, when a weak symmetry-breaking field h is introduced, the symmetry-breaking choice of each KZ domain is no longer completely random. Instead, the external field biases the local selection of the order parameter, making the process analogous to a biased coin toss that favors the symmetry-broken state correspondarXiv:2606.

p. 0
Where the paper says it

However, when a weak symmetry-breaking field h is introduced, the symmetry-breaking choice of each KZ domain is no longer completely random. Instead, the external field biases the local selection of the order parameter, making the process analogous to a biased coin toss that favors the symmetry-broken state correspondarXiv:2606.

27835#pg2> To test the theory quantitatively, we parameterize the defect density as n(d) defects = nKZ exp − λ τ(d+2)/4Q i, where λ is the suppression coefficient <ref:2606. p. 3
Where the paper says it

[8] for the 2D case). To test the theory quantitatively, we parameterize the defect density as ndefects = nKZ exph −λ τ (d+2)/4 Q i , where λ is the suppression coefficient.

27835#pg2> All data collapse onto a single master curve when λ is plotted as a function of the scaling variable h squared / (θ η(d+2)/4) <ref:2606. p. 3
Where the paper says it

As predicted by Eq. (5), all data collapse onto a single master curve when λ is plotted as a function of the scaling variable h 2/ θ η(d+2)/4 , as shown in Figs.

27835#pg2>.

Unified Large-Deviation Origin

The Z2 and U(1) descriptions reveal a unified large-deviation origin: the external field biases the order parameter, and defect formation requires a stochastic fluctuation capable of overcoming this bias throughout the characteristic mean-field freeze-out spacetime volume <ref:2606.27835#pg2> While a Z2 defect arises locally from a single domain choosing the disfavored minimum, a U(1) defect emerges from a collective phase mismatch along a multi-block plaquette <ref:2606. p. 9

Where the paper says it

(S33) Remarkably, the Z2 and U(1) descriptions reveal a unified large-deviation origin: the external field biases the order parameter, and defect formation requires a stochastic fluctuation capable of overcoming this bias throughout the characteristic mean-field freeze-out spacetime volume. While a Z2 defect arises locally from a single domain choosing the disfavored minimum, a U(1) defect emerges from a collective phase mismatch along a multi-block plaquette.

27835#pg2> Despite these distinct topological structures, both models collapse onto the exact same universal large-deviation scaling exponent dictated by the fluctuation-dissipation scale θ, verifying the consistency of the FW framework across different symmetry groups under Ginzburg–Landau dynamics <ref:2606. p. 9
Where the paper says it

While a Z2 defect arises locally from a single domain choosing the disfavored minimum, a U(1) defect emerges from a collective phase mismatch along a multi-block plaquette. Despite these distinct topological structures, both models collapse onto the exact same universal large-deviation scaling exponent dictated by the fluctuation-dissipation scale θ, verifying the consistency of the FW framework across different symmetry groups under Ginzburg–Landau dynamics.

27835#pg2>.

Conclusion and Outlook

The agreement between theory and simulations is excellent in both one and two dimensions, providing strong evidence that weak symmetry breaking transforms defect formation from a Kibble–Zurek process into a rare-event problem <ref:2606.27835#pg2> These findings are directly relevant to realistic phase transitions, where perfect symmetry is rarely realized in practice <ref:2606. p. 3

Where the paper says it

Outlook – The agreement between theory and simulations is excellent in both one and two dimensions, providing strong evidence that weak symmetry breaking transforms defect formation from a Kibble–Zurek process into a rare-event problem. These findings are directly relevant to realistic phase transitions, where perfect symmetry is rarely realized in practice.

p. 3
Where the paper says it

Outlook – The agreement between theory and simulations is excellent in both one and two dimensions, providing strong evidence that weak symmetry breaking transforms defect formation from a Kibble–Zurek process into a rare-event problem. These findings are directly relevant to realistic phase transitions, where perfect symmetry is rarely realized in practice.

27835#pg2> In particular, symmetry-violating biases have already been shown experimentally to suppress the Kibble–Zurek mechanism in quantum superfluids <ref:2606. p. 3
Where the paper says it

These findings are directly relevant to realistic phase transitions, where perfect symmetry is rarely realized in practice. In particular, symmetry-violating biases have already been shown experimentally to suppress the Kibble–Zurek mechanism in quantum superfluids [10], and analogous effects can be engineered in quantum phase transitions using external bias fields [11].

p. 3
Where the paper says it

These findings are directly relevant to realistic phase transitions, where perfect symmetry is rarely realized in practice. In particular, symmetry-violating biases have already been shown experimentally to suppress the Kibble–Zurek mechanism in quantum superfluids [10], and analogous effects can be engineered in quantum phase transitions using external bias fields [11].

27835#pg2> Our theory should therefore apply directly to these settings and be immediately testable in current experimental platforms <ref:2606. p. 3
Where the paper says it

In particular, symmetry-violating biases have already been shown experimentally to suppress the Kibble–Zurek mechanism in quantum superfluids [10], and analogous effects can be engineered in quantum phase transitions using external bias fields [11]. Our theory should therefore apply directly to these settings and be immediately testable in current experimental platforms.

p. 3
Where the paper says it

In particular, symmetry-violating biases have already been shown experimentally to suppress the Kibble–Zurek mechanism in quantum superfluids [10], and analogous effects can be engineered in quantum phase transitions using external bias fields [11]. Our theory should therefore apply directly to these settings and be immediately testable in current experimental platforms.

27835#pg2>. The present analysis is restricted to the mean-field approximation, but extending this framework to include beyond-mean-field corrections remains an appealing avenue for future work <ref:2606. p. 3
Where the paper says it

Our theory should therefore apply directly to these settings and be immediately testable in current experimental platforms. Finally, while the present analysis is restricted to the mean-field approximation, extending this framework to include beyond-mean-field corrections remains an appealing avenue for future work.

27835#pg2>.

Supplementary Material Details

The Freidlin–Wentzell action functional provides the path probability of a field history P[ϕ] ≍ exp [−SFW[ϕ]] <ref:2606.27835#pg2> For the Z2 model, the activation barrier scales as S(d) h ∼ h squared ηθ tˆ d ∼ h squared θ / (τQ η d+2/4) <ref:2606. p. 7

Where the paper says it

Rescaling time by t/tˆ shows that the remaining integrals contribute only order-one numerical constants. Therefore, in d spatial dimensions, the activation barrier scales as S (d) h ∼ h 2 ηθ tˆˆξ d ∼ h 2 θ τQ η d+2 4 .

27835#pg2> For the U(1) model in two dimensions, the explicit field contribution scales precisely as Sh ∼ A h squared 4ηθ tˆ 2 ∼ A h squared 4 h squared τQ η d+2/4 <ref:2606. p. 8
Where the paper says it

Evaluating this integral over the freeze-out temporal window and scaling the boundary as R ∼ ˆξ directly provides the activation barrier. Using the explicit mean-field solutions for tˆ and ˆξ to construct the characteristic spacetime volume tˆˆξ 2 ∼ τQ, we find that the explicit field contribution scales precisely as: Sh ∼ A2 h 2 4ηθ tˆˆξ 2 ∼ A2 4 h 2 τQ θη , (S32)

27835#pg2> The final defect density for the U(1) model is nv ∼ η τQ / 1/2 exp − A h squared 4 h squared τQ η d+2/4 <ref:2606. p. 9
Where the paper says it

10 where A2 is an order-one constant collecting the spatial core cutoffs, stiffness parameters, and profile details. Restoring this to the density relation (S26) yields the final defect density verified in the main text: nv ∼ η τQ 1/2 exp − A2 4 h 2 τQ θη .

27835#pg2>. The numerical simulations track defects as sign changes of the order parameter between adjacent lattice points, yielding Nkink = N x(x-1) i=1 1 − σiσi+1 / 2 <ref:2606. p. 9
Where the paper says it

As the quench progresses deep into the broken-symmetry regime, distinct spatial domains settle into opposite signs of ϕ; the zeros separating these domains constitute the one-dimensional Z2 defects. Defects in the one-dimensional Z2 simulations are tracked as sign changes of the order parameter between adjacent lattice points.

27835#pg2> The total number of vortices is measured at a late characteristic time scale t = ctˆ (where c is an order-one constant) <ref:2606. p. 9
Where the paper says it

The total defect number is subsequently recorded at a late time: tcount = 2τQ. (S4) This delayed counting procedure eliminates ambiguities associated with the frozen near-critical regime and measures the topological defects only after the order parameter has firmly settled into the broken-symmetry basins.

27835#pg2>.

Appendix S1 and S2 Summary

In the Z2 model, the path probability scales as P[ϕ] ≍ exp [−SFW[ϕ]] where SFW[ϕ] = 1 / (4ηθ) Z tˆ - tˆ dt ddx η∂tϕ − ∇2ϕ + a 3 ϕ - b(t) ϕ - h squared <ref:2606.27835#pg2> The action cost Sh is derived as Sblock[ϕ] = ˆξ d 4ηθ Z tˆ - tˆ dt η∂tϕ − t/τQ ϕ − h squared <ref:2606. p. 6

Where the paper says it

Setting ϕ(x, t) ≃ ϕ(t) eliminates the gradient term in this block approximation. Neglecting the cubic term along the early, near-critical segment of the path gives Sblock[ϕ] = ˆξ d 4ηθ Z tˆ −tˆ dt ηϕ˙ − t τQ ϕ − h 2 .

27835#pg2> In the U(1) model, the FW action is SFW[ψ] = 1 / (4ηθ) Z dt d2x η∂tψ + a(t)ψ + bψ squared ψ - γ∇ 2ψ - h squared <ref:2606. p. 8
Where the paper says it

The resulting vortex density scales as: nv ∼ ˆξ −2 pv(h, θ) ∼ η τQ 1/2 pv(h, θ). (S26) To evaluate the plaquette winding probability pv, we formulate the Freidlin–Wentzell (FW) action functional for the complex field dynamics: SFW[ψ] = 1 4ηθ Z dt d2x η∂tψ + a(t)ψ + b|ψ| 2ψ − γ∇2ψ − h 2 .

27835#pg2> The cross term between the winding phase factor e imφ and the uniform source h vanishes identically upon angular integration <ref:2606. p. 8
Where the paper says it

(S29) into the FW action leads to: SFW ≃ 1 4ηθ Z tˆ −tˆ dt Z R ε r dr Z 2π 0 dφ e imφ ηψ˙ + a(t)ψ + γm2ψ r 2 − h 2 . (S30) For elementary vortices (m = ±1), the cross term between the winding phase factor e imφ and the uniform source h vanishes identically upon angular integration.

27835#pg2>. The explicit symmetry-breaking contribution isolates cleanly into a distinct energy penalty Sh ≃ 1 / (4ηθ) Z tˆ - tˆ dt π(R squared − ε 2)h squared <ref:2606. p. 8
Where the paper says it

(S30) For elementary vortices (m = ±1), the cross term between the winding phase factor e imφ and the uniform source h vanishes identically upon angular integration. Consequently, the explicit symmetry-breaking contribution isolates cleanly into a distinct energy penalty: Sh ≃ 1 4ηθ Z tˆ −tˆ dt π(R 2 − ε 2 )h 2 , (S31) which physically represents the energetic cost required to maintain a topologically trapped winding configuration that cannot globally align with the external field.

27835#pg2> This leads to the final defect density nv ∼ η τQ / 1/2 exp − A h squared 4 h squared τQ η d+2/4 <ref:2606. p. 9
Where the paper says it

10 where A2 is an order-one constant collecting the spatial core cutoffs, stiffness parameters, and profile details. Restoring this to the density relation (S26) yields the final defect density verified in the main text: nv ∼ η τQ 1/2 exp − A2 4 h 2 τQ θη .

27835#pg2>. The numerical methods involve solving the overdamped stochastic differential equation using a second-order stochastic Heun predictor-corrector scheme <ref:2606. p. 9
Where the paper says it

(S2) On a discrete lattice with spatial spacing dx and temporal step dt, the stochastic increment is sampled with variance 2ηθ dt/dx. The spatial Laplacian is evaluated pseudospectrally via Fast Fourier Transforms (FFTs), and the time evolution is integrated using a second-order stochastic Heun predictor-corrector scheme.

27835#pg2> and counting defects as sign changes of the order parameter between adjacent lattice points <ref:2606. p. 9
Where the paper says it

As the quench progresses deep into the broken-symmetry regime, distinct spatial domains settle into opposite signs of ϕ; the zeros separating these domains constitute the one-dimensional Z2 defects. Defects in the one-dimensional Z2 simulations are tracked as sign changes of the order parameter between adjacent lattice points.

27835#pg2>. The simulation setup involves a system size L = 2048, Nx = 16384, dx = L / Nx, with a fixed time step of dt = 0. p. 9
Where the paper says it

The spatial Laplacian is evaluated pseudospectrally via Fast Fourier Transforms (FFTs), and the time evolution is integrated using a second-order stochastic Heun predictor-corrector scheme. The system size and grid dimensions utilized in the production runs are: L = 2048, Nx = 16384, dx = L Nx , (S3) with a fixed time step of dt = 0.

1 <ref:2606. p. 0
Where the paper says it

However, when a weak symmetry-breaking field h is introduced, the symmetry-breaking choice of each KZ domain is no longer completely random. Instead, the external field biases the local selection of the order parameter, making the process analogous to a biased coin toss that favors the symmetry-broken state correspondarXiv:2606.

27835#pg2> and the total defect number is recorded at a late time tcount = 2τQ <ref:2606. p. 9
Where the paper says it

In the simulations, the field is evolved well past the transition point to ensure that independent domains and sharp kinks are fully developed. The total defect number is subsequently recorded at a late time: tcount = 2τQ.

27835#pg2>. The U(1) simulations use a periodic box of size L = 500 discretized on a square grid of Nx × Ny = 500 × 500 points (dx = 1) <ref:2606.27835#pg2>. The vortices are identified by evaluating the discrete phase winding over elementary four-point plaquettes, calculating the contour integral of the phase gradient to identify singularities where W = ±1 <ref:2606. p. 8
Where the paper says it

(S29) into the FW action leads to: SFW ≃ 1 4ηθ Z tˆ −tˆ dt Z R ε r dr Z 2π 0 dφ e imφ ηψ˙ + a(t)ψ + γm2ψ r 2 − h 2 . (S30) For elementary vortices (m = ±1), the cross term between the winding phase factor e imφ and the uniform source h vanishes identically upon angular integration.

27835#pg2> and local subdomains of the form ψ i = ψ ie iφ i are well-formed at a late time t = cτˆ (where c is an order-one constant) <ref:2606. p. 9
Where the paper says it

In the simulations, the field is evolved well past the transition point to ensure that independent domains and sharp kinks are fully developed. The total defect number is subsequently recorded at a late time: tcount = 2τQ.

27835#pg2>. The ensemble average of Nkink over all independent stochastic realizations yields the reported defect number for each parameter set <ref:2606. p. 9
Where the paper says it

The system size and grid dimensions utilized in the production runs are: L = 2048, Nx = 16384, dx = L Nx , (S3) with a fixed time step of dt = 0. For each parameter set, the dynamics are evolved over 20 independent noise realizations in parallel.

27835#pg2>. The text emphasizes that the derivation of Eq. p. 5
Where the paper says it

6 Supplementary Material In this Supplementary Material, we provide a formal re-derivation of our main result, Eq. (5) of the main text, using large-deviation theory and, in particular, the Freidlin–Wentzell action functional within the mean-field approximation.

(5) is fundamentally different from the nucleation-theory corrections arising near weakly first-order transitions <ref:2606. p. 2
Where the paper says it

We emphasize that the derivation of Eq. (5) is fundamentally different from the nucleation-theory corrections arising near weakly first-order transitions [7].

27835#pg2>.

Summary of Key Findings

The research demonstrates that weak symmetry breaking transforms defect formation from a Kibble–Zurek process into a rare-event problem <ref:2606.27835#pg2> The key result is the closed-form expression for the defect density in arbitrary dimensions, valid in the weak-field and weak-noise limits <ref:2606. p. 0

Where the paper says it

Here, using large-deviation theory, we show that defect formation under weak symmetry breaking is controlled by rare fluctuations that drive local regions into the disfavored symmetry-broken state. This mechanism yields a closed-form expression for the defect density in arbitrary dimensions, valid in the weak-field and weak-noise limits.

27835#pg2> This result is controlled by rare fluctuations that coherently bias an entire Kibble–Zurek domain toward the disfavored vacuum during freeze-out <ref:2606. p. 2
Where the paper says it

In the latter case, rare events are associated with the nucleation of critical droplets that overcome a free-energy barrier. By contrast, in the present scenario defect formation is controlled by rare fluctuations that coherently bias an entire Kibble–Zurek domain toward the disfavored vacuum during freeze-out, without requiring activated crossing of the potential barrier ∆V shown in Fig.

27835#pg2> The final defect density is given by n(d) rare ∼ η τQ d 4 exp [−Ad h squared / (θ τQ η d+2/4)] <ref:2606. p. 9
Where the paper says it

10 where A2 is an order-one constant collecting the spatial core cutoffs, stiffness parameters, and profile details. Restoring this to the density relation (S26) yields the final defect density verified in the main text: nv ∼ η τQ 1/2 exp − A2 4 h 2 τQ θη .

27835#pg2>. This result is verified numerically in both one and two dimensions, showing a clear exponential correction to the standard KZ power-law scaling <ref:2606. p. 3
Where the paper says it

We numerically compute the defect density generated during nonequilibrium quenches and determine its scaling behavior in the slow-quench regime by averaging over many independent realizations. In both one and two dimensions, we observe a clear exponential correction to the standard KZ power-law scaling, consistent with the numerical findings of Ref.

27835#pg2> The theory unifies the description of Z2 and U(1) topological defects under Ginzburg–Landau dynamics <ref:2606. p. 9
Where the paper says it

While a Z2 defect arises locally from a single domain choosing the disfavored minimum, a U(1) defect emerges from a collective phase mismatch along a multi-block plaquette. Despite these distinct topological structures, both models collapse onto the exact same universal large-deviation scaling exponent dictated by the fluctuation-dissipation scale θ, verifying the consistency of the FW framework across different symmetry groups under Ginzburg–Landau dynamics.

27835#pg2>.

Outlook for Future Work

The agreement between theory and simulations is excellent in both one and two dimensions, providing strong evidence that weak symmetry breaking transforms defect formation from a Kibble–Zurek process into a rare-event problem <ref:2606.27835#pg2> These findings are directly relevant to realistic phase transitions, where perfect symmetry is rarely realized in practice <ref:2606.27835#pg2> In particular, symmetry-violating biases have already been shown experimentally to suppress the Kibble–Zurek mechanism in quantum superfluids <ref:2606.27835#pg2> Our theory should therefore apply directly to these settings and be immediately testable in current experimental platforms <ref:2606.27835#pg2> Finally, while the present analysis is restricted to the mean-field approximation, extending this framework to include beyond-mean-field corrections remains an appealing avenue for future work <ref:2606. p. 3

Where the paper says it

Our theory should therefore apply directly to these settings and be immediately testable in current experimental platforms. Finally, while the present analysis is restricted to the mean-field approximation, extending this framework to include beyond-mean-field corrections remains an appealing avenue for future work.

27835#pg2>. The authors acknowledge support from H. p. 3
Where the paper says it

[10] to us. JL and MB acknowledge the support of the Shanghai Municipal Science and Technology Major Project (Grant No.

B. Zeng and G. p. 3
Where the paper says it

We thank H. Zeng for useful comments and G.

Volovik <ref:2606. p. 3
Where the paper says it

Zeng for useful comments and G. Volovik for pointing our Ref.

27835#pg2> and the Shanghai Municipal Science and Technology Major Project <ref:2606. p. 3
Where the paper says it

[10] to us. JL and MB acknowledge the support of the Shanghai Municipal Science and Technology Major Project (Grant No.

27835#pg2>. The paper is available as arXiv:2606. p. 0
Where the paper says it

However, when a weak symmetry-breaking field h is introduced, the symmetry-breaking choice of each KZ domain is no longer completely random. Instead, the external field biases the local selection of the order parameter, making the process analogous to a biased coin toss that favors the symmetry-broken state correspondarXiv:2606.

27835v2 [cond-mat. p. 0
Where the paper says it

Instead, the external field biases the local selection of the order parameter, making the process analogous to a biased coin toss that favors the symmetry-broken state correspondarXiv:2606. 27835v2 [cond-mat.

stat-mech] 15 Jul 2026 <ref:2606. p. 0
Where the paper says it

27835v2 [cond-mat. stat-mech] 15 Jul 2026

27835#pg2>

--- Page 1 ---

Rare Events Govern Defect Formation under Weak Symmetry Breaking

Jiang Liu, Peng Yang, Matteo Baggioli <ref:2606.27835#pg2> Wilczek Quantum Center, School of Physics and Astronomy, Shanghai Jiao Tong University, Shanghai 200240, China 2Shanghai Research Center for Quantum Sciences, Shanghai 201315,China 3Department of Physics, The Chinese University of Hong Kong, Shatin, New Territories, Hong Kong 999077 China and 4The state Key Laboratory of Quantum Information Technologies and Materials, The Chinese University of Hong Kong Shatin New Territories Hong Kong 999077 China

--- Page 2 ---

2

FIG. 1: Weak Symmetry Breaking as a Biased Coin Toss. p. 0

Where the paper says it

However, when a weak symmetry-breaking field h is introduced, the symmetry-breaking choice of each KZ domain is no longer completely random. Instead, the external field biases the local selection of the order parameter, making the process analogous to a biased coin toss that favors the symmetry-broken state correspondarXiv:2606.

Schematic illustration for the simple case of a Z2-symmetric system. p. 1
Where the paper says it

1: Weak Symmetry Breaking as a Biased Coin Toss. Schematic illustration for the simple case of a Z2-symmetric system.

(a) In the absence of explicit symmetry breaking (h = 0), independent Kibble–Zurek domains of size ξ choose between the two degenerate symmetry-broken states with equal probability, P+ = P− = 1/2 <ref:2606. p. 1
Where the paper says it

Schematic illustration for the simple case of a Z2-symmetric system. (a) In the absence of explicit symmetry breaking (h = 0), independent Kibble–Zurek domains of size ˆξ choose between the two degenerate symmetry-broken states with equal probability, P+ = P− = 1/2.

27835#pg2> (b) A weak external field h biases the local symmetry-breaking choice, leading to unequal probabilities P+ > P− <ref:2606. p. 1
Where the paper says it

(a) In the absence of explicit symmetry breaking (h = 0), independent Kibble–Zurek domains of size ˆξ choose between the two degenerate symmetry-broken states with equal probability, P+ = P− = 1/2. (b) A weak external field h biases the local symmetry-breaking choice, leading to unequal probabilities P+ > P−.

27835#pg2> (c) Defect formation requires a rare fluctuation, occurring with probability Pwrong, that drives a Kibble–Zurek domain into the disfavored metastable minimum against the bias imposed by the external field <ref:2606. p. 1
Where the paper says it

(b) A weak external field h biases the local symmetry-breaking choice, leading to unequal probabilities P+ > P−. (c) Defect formation requires a rare fluctuation, occurring with probability Pwrong, that drives a Kibble–Zurek domain into the disfavored metastable minimum against the bias imposed by the external field.

27835#pg2> (d) Pwrong corresponds to the probability of rare fluctuations in the noise distribution that are sufficiently strong to overcome the bias introduced by the external field h <ref:2606. p. 1
Where the paper says it

(c) Defect formation requires a rare fluctuation, occurring with probability Pwrong, that drives a Kibble–Zurek domain into the disfavored metastable minimum against the bias imposed by the external field. (d) Pwrong corresponds to the probability of rare fluctuations in the noise distribution that are sufficiently strong to overcome the bias introduced by the external field h.

27835#pg2>

--- Page 3 ---

3

(a) 1D h = 0 h > 0 t/τQ (b) Defect density in the 1D Ginzburg–Landau model with and without an external symmetry-breaking field h. The black curve shows the standard Kibble–Zurek power-law scaling, while the red curve corresponds to the theoretical prediction including the exponential correction of Eq. p. 2

Where the paper says it

(b) Defect density in the 1D Ginzburg–Landau model with and without an external symmetry-breaking field h. The black curve shows the standard Kibble–Zurek power-law scaling, while the red curve corresponds to the theoretical prediction including the exponential correction of Eq.

(5) <ref:2606.27835#pg2> (c) Action cost Sh as a function of τQ for the 1D model at different values of h. p. 2
Where the paper says it

The black curve shows the standard Kibble–Zurek power-law scaling, while the red curve corresponds to the theoretical prediction including the exponential correction of Eq. (c) Action cost Sh as a function of τQ for the 1D model at different values of the external field h.

Solid lines are fits to the predicted τ(3/4Q) scaling of Eq. p. 2
Where the paper says it

(c) Action cost Sh as a function of τQ for the 1D model at different values of the external field h. Solid lines are fits to the predicted τ 3/4 Q scaling of Eq.

(5) <ref:2606.27835#pg2>

--- Page 4 ---

4

(d,e) Universal collapse of the parameter λ onto the scaling form predicted by Eq. (5) for both the 1D and 2D systems <ref:2606. p. 0

Where the paper says it

However, when a weak symmetry-breaking field h is introduced, the symmetry-breaking choice of each KZ domain is no longer completely random. Instead, the external field biases the local selection of the order parameter, making the process analogous to a biased coin toss that favors the symmetry-broken state correspondarXiv:2606.

27835#pg2>

--- Page 5 ---

5

n(d) defects ∼ η τQ d 4 exp [−Ad h squared / (θ τQ η d+2/4)] <ref:2606.27835#pg2>

--- Page 6 ---

6

Supplementary Material In this Supplementary Material, we provide a formal re-derivation of our main result, Eq. (5) of the main text, using large-deviation theory and, in particular, the Freidlin–Wentzell action functional within the mean-field approximation <ref:2606. p. 5

Where the paper says it

6 Supplementary Material In this Supplementary Material, we provide a formal re-derivation of our main result, Eq. (5) of the main text, using large-deviation theory and, in particular, the Freidlin–Wentzell action functional within the mean-field approximation.

27835#pg2> In this Supplementary Material, we provide a formal re-derivation of our main result, Eq. p. 5
Where the paper says it

6 Supplementary Material In this Supplementary Material, we provide a formal re-derivation of our main result, Eq. (5) of the main text, using large-deviation theory and, in particular, the Freidlin–Wentzell action functional within the mean-field approximation.

p. 5
Where the paper says it

6 Supplementary Material In this Supplementary Material, we provide a formal re-derivation of our main result, Eq. (5) of the main text, using large-deviation theory and, in particular, the Freidlin–Wentzell action functional within the mean-field approximation.

(5) of the main text using large-deviation theory and in particular the Freidlin–Wentzell action functional within the mean-field approximation <ref:2606. p. 5
Where the paper says it

6 Supplementary Material In this Supplementary Material, we provide a formal re-derivation of our main result, Eq. (5) of the main text, using large-deviation theory and, in particular, the Freidlin–Wentzell action functional within the mean-field approximation.

p. 5
Where the paper says it

6 Supplementary Material In this Supplementary Material, we provide a formal re-derivation of our main result, Eq. (5) of the main text, using large-deviation theory and, in particular, the Freidlin–Wentzell action functional within the mean-field approximation.

27835#pg2> In this Supplementary Material, we provide a formal re-derivation of our main result, Eq. (5) of the main text using large-deviation theory and in particular the Freidlin–Wentzell action functional within the mean-field approximation <ref:2606.27835#pg2>

--- Page 7 ---

7

SM1. Freidlin–Wentzell action functional In the weak additive-noise limit, fluctuations of a stochastically perturbed dynamical system are governed, at leading exponential order, by Large Deviation Theory (LDT) [9, 12] <ref:2606. p. 5

Where the paper says it

Appendix S1: Large deviation theory SM1. Freidlin–Wentzell action functional In the weak additive-noise limit, fluctuations of a stochastically perturbed dynamical system are governed, at leading exponential order, by Large Deviation Theory (LDT) [9, 12].

27835#pg2> Thus, the action is the quadratic cost of the noise realization needed to make the system follow the path x(t): deterministic trajectories have ˙x = a(x), while deviations from this drift are exponentially suppressed <ref:2606. p. 5
Where the paper says it

For a finite-dimensional additive-noise process X˙ = a(X) + ϵW , ˙ the FW estimate for a path x(t) over the time interval [0, T] reads P[X(t) ≃ x(t)] ≍ exp " − 1 2ϵ 2 Z T 0 dt ∥x˙ − a(x)∥ 2 # , where the symbol ≍ signifies asymptotic equivalence at leading exponential order. Thus, the action is the quadratic cost of the noise realization needed to make the system follow the path x(t): deterministic trajectories have ˙x = a(x), while deviations from this drift are exponentially suppressed.

27835#pg2> For our concrete purpose, we start from the overdamped Langevin equation for a real scalar field ϕ(x, t), η∂tϕ = − δF / δϕ + ξ(x, t) <ref:2606. p. 5
Where the paper says it

For the field theory below, the same idea gives the path probability of a field history ϕ(x, t); the only change is that the cost is integrated over both time and space. For our concrete purpose, we start from the overdamped Langevin equation for a real scalar field ϕ(x, t), η∂tϕ = − δF δϕ + ξ(x, t), (S1) with the noise correlation given by ⟨ξ(x, t)ξ(x ′ , t′ )⟩ = 2ηθ δ(d) (x − x ′ )δ(t − t ′ ).

27835#pg2> The path probability therefore has the asymptotic form P[ϕ] ≍ exp [−SFW[ϕ]] <ref:2606. p. 5
Where the paper says it

The deterministic drift is a[ϕ] = − 1 η δF δϕ , so the mismatch ∂tϕ − a[ϕ] is the noise history required to generate the path ϕ(x, t). The path probability therefore has the asymptotic form P[ϕ] ≍ exp [−SFW[ϕ]] , (S4) with SFW[ϕ] = 1 2ϵ 2 Z dt ddx ∂tϕ + 1 η δF δϕ 2 = 1 4ηθ Z dt ddx η∂tϕ + δF δϕ 2 .

27835#pg2>

--- Page 8 ---

8

SM1. Freidlin–Wentzell action functional SM2. p. 2

Where the paper says it

Importantly, Eq. (5) can also be derived from a formal large-deviation analysis based on minimizing the Freidlin–Wentzell action functional [9], which yields the same result.

Z2 model We consider a real scalar order parameter ϕ(x, t) governed by overdamped stochastic dynamics with Ginzburg–Landau-type evolution: η∂tϕ(x, t) − ∇ 2ϕ(x, t) + ∂ϕV (ϕ, t) = θ(x, t), (S6) where η is the damping coefficient <ref:2606. p. 6
Where the paper says it

Z2 model We consider a real scalar order parameter ϕ(x, t) governed by overdamped stochastic dynamics with Ginzburg– Landau-type evolution: η∂tϕ(x, t) − ∇2ϕ(x, t) + ∂ϕV (ϕ, t) = ϑ(x, t), (S6) where η is the damping coefficient. The time-dependent potential is V (ϕ, t) = a 4 ϕ 4 − b(t) 2 ϕ 2 − hϕ, a > 0, b(t) = t τQ , (S7) where τQ is the quench time and h > 0 favors the positive phase.

27835#pg2> The Kibble–Zurek condition τrel(tˆ) = tˆ then yields tˆ=√ητQ, ˆξ = τQ / η 1/4 <ref:2606. p. 6
Where the paper says it

The relaxation time and correlation length on the stable side are τrel(t) = η |b(t)| = ητQ |t| , ξ(t) = 1 p |b(t)| = rτQ |t| . (S9) The Kibble–Zurek condition τrel(tˆ) = tˆ then yields tˆ= √ ητQ, ˆξ = τQ η 1/4 .

27835#pg2> Thus, this explicit Ginzburg–Landau calculation yields the mean-field exponents of the Gaussian Model-A: νMF = 1/2 and z MF = 2 <ref:2606. p. 6
Where the paper says it

(S9) The Kibble–Zurek condition τrel(tˆ) = tˆ then yields tˆ= √ ητQ, ˆξ = τQ η 1/4 . (S10) Thus, this explicit Ginzburg–Landau calculation yields the mean-field exponents of the Gaussian Model-A: νMF = 1/2 and zMF = 2.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Integration of rare event theory into defect formation models will allow AI systems to accurately predict topological defect densities in non-equilibrium systems undergoing weak symmetry breaking. This enables AI to move beyond standard Kibble-Zurek scaling to capture exponential corrections in realistic physical scenarios, such as quantum phase transitions or materials synthesis simulations.

  2. Development of a framework based on the Freidlin Wentzell action functional will allow AI to calculate the probability of rare fluctuations that drive local regions into disfavored symmetry-broken states. This capability will enable AI to identify and predict metastable configurations in complex physical systems where defects are governed by overcoming energy barriers, rather than simple random mismatch.

  3. Application of the derived scaling relation n(d)rare η τQ d 4 exp [-Ad h squared / (θ τQ η d+2)] will allow AI to quantify the suppression of defect formation as a function of external bias field h and noise strength theta. This will enable AI to design experiments or simulate materials where defects are suppressed by specific physical parameters, leading to more robust material design predictions.

  4. The unified large-deviation origin revealed by comparing Z2 and U1 models will allow AI to generalize defect physics across different symmetry groups in Ginzburg-Landau dynamics. This generalization enables AI models to handle a wider variety of physical systems, from simple Ising models to complex quantum phase transitions, with a single theoretical structure for predicting topological features.

Abstract

Crossing a continuous phase transition out of equilibrium typically generates topological defects whose density obeys a universal power-law scaling predicted by the Kibble-Zurek mechanism. Recent numerical studies have revealed systematic deviations from this scaling in the presence of weak explicit symmetry breaking, manifested as an additional exponential suppression of defect formation. However, the origin of this correction and a general theoretical framework to describe it have remained elusive. Here, using large-deviation theory, we show that defect formation under weak symmetry breaking is controlled by rare fluctuations that drive local regions into the disfavored symmetry-broken state. Within mean-field theory, this mechanism yields a closed-form expression for the defect density in arbitrary dimensions, valid in the weak-field and weak-noise limits. These theoretical predictions are verified through direct simulations of stochastic Ginzburg-Landau models in one and two spatial dimensions.

Related papers