Quantum annealing through a first-order phase transition: Field theory approach

arXiv:2608.21329 · quant-ph, cond-mat.other · Submitted 2026-08-21 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Quantum Radio. Generated commentary on the latest quantum physics and condensed matter papers.

Kai: We're looking at the paper: "Quantum annealing through a first-order phase transition".

Mira: Here's the gist of the paper. Quantum annealing through a first-order phase transition: field theory approach is a field theory approach that predicts several power laws for the error generation rate during quantum annealing either though or near the first-order phase transition.

Kai: The Core Concept of First-Order Transitions Unlike second-order phase transitions, a first-order transition has a stage, in which a system is trapped in a metastable state.

Mira: The decay of this state leads to abundant excitations over the ground state.

Kai: This mechanism allows defects at a first-order transition to be produced even in the adiabatic limit when the system is trapped for some time in a false vacuum, i.e., a non-global energy minimum.

Title and authors: Kai: Let's talk about this paper, "Quantum annealing through a first-order phase transition: Field theory approach," and who wrote it. It’s by Diego Andr´es Rivera Orona, Predrag Punoˇsevac, and Nikolai A. Sinitsyn.

Mira: The title itself tells us the core idea is using a field theory to look at quantum annealing through the lens of a first-order phase transition instead of just treating it as something else.

Lev: And this approach is interesting because it connects directly to how we model these systems, specifically by using the driven Lipkin-Meshkov-Glick model, or LMGm, as the minimal interacting Ising spins.

Kai: So they use that LMGm to demonstrate this first-order behavior, which sets the stage for applying their field theory description to actual quantum annealing problems.

Mira: The paper sets up a Lagrangian for this field near the critical point using a real scalar field with an external force, and they define the potential energy by the expression V of phi t equals minus beta t phi minus k phi squared over two plus g phi to the power four over four.

Lev: That potential setup is crucial because it allows them to analyze how many errors emerge during quantum annealing computations when time and beta are involved in that evolution.

Kai: And they show that the first-order critical point itself emerges when the quadratic term in that equation is negative, meaning k has to be greater than zero.

Mira: That condition on k is what really separates the regimes, because it dictates whether we're looking at a transition where a first-order phase transition actually happens or not.

Lev: So this framework helps us define the parameter space for these quantum annealing problems based on those field theory parameters m, k, and g.

Kai: And they connect these field theory parameters to the spin Hamiltonian through relationships like m equals one over Bx S, k equals D minus Bx S, and g equals Bx two over S cubed.

Mira: Those relationships show us how the microscopic details of our spin system translate into the macroscopic description of that scalar field.

Lev: It’s a useful bridge because we can take something abstract from the field theory and connect it to something concrete in terms of our spin Hamiltonian parameters for testing.

The paper's summary: Kai: To summarize, the paper is presenting this field theory for kinetics of defects emerging during quantum annealing computations. It predicts several power laws for the error generation rate either through or near the first-order phase transition.

Mira: They focus on how these power laws depend on whether k is less than zero, equal to zero, or greater than zero, which tells us a lot about the behavior in different regimes.

Lev: Specifically for k less than zero, they predict nex scales with beta times a minus delta over beta to the power of nu minus where delta is proportional to the absolute value of k times mu minus >

Kai: For when k equals zero, they predict nex scales with beta times nu for some exponent nu which is shown in equation seven.

Mira: And for when k greater than zero, they predict nex scales with an initial number n oh plus c times beta times nu plus where n oh doesn't depend on beta unless you take into account exponentially suppressed quantum over-barrier tunneling leading to extremely slow decay of that term.

Lev: So the paper is essentially mapping out how the error rate changes based on those different regimes of k, which is key for understanding the kinetics.

Kai: It’s about describing what happens when you have that metastable state decaying into abundant excitations over the ground state because of a first-order transition.

Mira: And they show that this mechanism allows defects at a first-order transition to be produced even in the adiabatic limit when you're trapped for some time in a false vacuum, or non-global energy minimum.

Lev: That means when that minimum disappears, the potential energy releases into large collective oscillations which then disperse into computational errors.

Kai: The LMGm model is just the minimal setup that proves this behavior and it shows exactly how those defects are produced in this context.

Mira: So the main point is that they provide a concrete model showing the mechanism of defect generation through these phase transitions, which is what we need to see.

The paper's improvements: Kai: The authors suggest a few things that could be improvements for this theory. One big suggestion is searching for sharp changes between the power exponents as you vary continuous parameters, which they call this a signature of first-order critical points.

Mira: That’s a way to detect those transition points experimentally or numerically, which would be a signature of first-order critical points and could help identify and avoid them for better computations.

Lev: From an engineering standpoint, it means we can use that to determine the necessary suppression of parameter uncertainties below a certain critical value based on the discussion in section four.

Kai: So if you have well-mixed interactions, this theory suggests we can actively suppress those parameter uncertainties below a certain level to avoid unwanted first-order critical points.

Mira: This is how we move from just observing the physics to actually designing more robust computations by using the theory to guide our engineering choices.

Lev: That moves us toward building systems that are less sensitive to those kinds of kinds of dynamical instabilities, which is a good direction for error correction research.

Conclusion: Kai: So, this paper, "Quantum annealing through a first-order phase transition: Field theory approach," really connects the physics of these transitions to the actual performance we see in quantum hardware.

Mira: It shows that the different scaling behaviors of defects correspond to different convergence rates of computational algorithms.

Lev: It's about using this field theory to understand how we can guide us toward better designs by identifying where those critical points are hiding in the parameter space.

Kai: The core message is that if you use these transition scalings, you can infer new algorithmic possibilities for some known problems based on these transition scalings.

Mira: It's a solid paper that gives us a theoretical framework to study how we can design more robust quantum annealing experiments with better control over the physics.

Lev: I just think it’s solid work that lays down the necessary foundation for figuring out the real engineering challenges we face in this space.

Diego Andrés Rivera Orona, Predrag Punošsevac, Nikolai A. Sinitsyn

Los Alamos National Laboratory

quant-ph, cond-mat.other

Submitted: 2026-08-21

Updated: 2026-11-16

Comments: 10 pages, 3 figures

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

Importance score: 79/100

The gist: Quantum annealing through a first-order phase transition: field theory approach is a field theory approach that predicts several power laws for the error generation rate during quantum annealing…

Key concepts

First-Order Phase Transition
Unlike smooth transitions, this involves a system getting stuck in a temporary state (metastable state) before suddenly releasing energy. This trapping mechanism allows errors to be created even when the process seems slow, as the trapped minimum eventually disappears.
Field Theory Formulation
The physics of the first-order transition is modeled using a real scalar field theory with an external force. The Lagrangian describes how this field behaves near the critical point, allowing researchers to derive specific mathematical relationships governing excitation rates.
Scaling Laws for Excitations
The field theory predicts different power laws for the number of excitations (errors) based on the distance 'k' from the transition boundary. These laws dictate how errors scale with transition rates and other system parameters, helping characterize different annealing regimes.

Terminology

Summary

Quantum annealing through a first-order phase transition: field theory approach is a field theory approach that predicts several power laws for the error generation rate during quantum annealing either though or near the first-order phase transition. p. 0

Where the paper says it

We present a field theory for kinetics of defects emerging during quantum annealing computations. This theory predicts several power laws for the error generation rate during quantum annealing either though or near the first-order phase transition.

The Core Concept of First-Order Transitions

Unlike second-order phase transitions, a first-order transition has a stage, in which a system is trapped in a metastable state. The decay of this state leads to abundant excitations over the ground state. p. 0

Where the paper says it

Sinitsyn1 1Theoretical Division, Los Alamos National Laboratory, Los Alamos, New Mexico 87545, USA 2 ISR Division, Los Alamos National Laboratory, Los Alamos, New Mexico 87545, USA Unlike second-order phase transitions, a first-order transition has a stage, in which a system is trapped in a metastable state. The decay of this state leads to abundant excitations over the ground state.

This mechanism allows defects at a first-order transition to be produced even in the adiabatic limit when the system is trapped for some time in a false vacuum, i. p. 0
Where the paper says it

How can excitations persist during very slow annealing? First-order quantum phase transitions provide a mechanism: The system is trapped for some time in a false vacuum, i.

e., a non-global energy minimum. p. 0
Where the paper says it

First-order quantum phase transitions provide a mechanism: The system is trapped for some time in a false vacuum, i. , a non-global energy minimum, as illustrated in Fig.

When this minimum disappears, the potential energy is released into large collective oscillations, which later disperse into computational errors. p. 0
Where the paper says it

1(c)), the potential energy is released into large collective oscillations, as in Fig. 1(d), which later disperse into computational errors [2].

The driven Lipkin-Meshkov-Glick model (LMGm) serves as the minimal model of interacting Ising spins that demonstrates this behavior. p. 0
Where the paper says it

Observation of this behavior would be a signature of first-order critical points and could help identify and avoid them for better computations. The driven Lipkin-Meshkov-Glick model (LMGm) serves as the minimal model of interacting Ising spins that demonstrates this behavior.

Field Theory Formulation

The minimal model of the first-order phase transition is described by a real scalar field theory with an external force. The Lagrangian of this field near the critical point is given by L = mϕ˙2 / 2 − V (ϕ, t). p. 0

Where the paper says it

Thus, the minimal model of the first-order phase transition is a real scalar ϕ 4 theory with an external force. The Lagrangian of this field near the critical point is L = mϕ˙2 2 − V (ϕ, t), (4) arXiv:2608.

The potential energy is defined as V (ϕ, t) = −βtϕ − kϕ2 / 2 + gϕ4 / 4. p. 1
Where the paper says it

Near t = 0, many levels pass very close to the ground-state energy, thus enhancing nonadiabatic behavior. where V (ϕ, t) = −βtϕ − kϕ2 2 + gϕ4 4 ; m, g, β > 0, (5) where t is time and β is the sweep rate of the crossing through the critical point during quantum annealing.

The first-order critical point emerges when the quadratic term in Eq. p. 1
Where the paper says it

Furthermore, m, k, and g are parameters that can be derived from a more specific quantum mechanical model near the critical point, as we will illustrate in Methods for the LMGm. The first-order critical point emerges when the quadratic term in Eq.

(5) is negative, i. p. 1
Where the paper says it

The first-order critical point emerges when the quadratic term in Eq. (5) is negative, i.

e., k > 0.

Scaling Laws for Excitations

The field theory predicts a set of power laws for the number of excitations, nex. These scaling laws describe the dependence of nex not only on the transition rate β but also on the characteristic distance k from the boundary between different quantum annealing regimes. p. 1

Where the paper says it

(4), are a set of power laws for the number of excitations. They describe the dependence of nex not only on the transition rate β but also on the characteristic distance k from the boundary between different quantum annealing regimes.

The predicted power laws are:

  1. For k < 0: nex ∝ β ν− − e − ∆/β, where ∆ ∝ k µ−. p. 0

    Where the paper says it

    The rate of defect production during a second-order phase transition is relatively well understood due to representative solvable models [2–8] and the phenomenological Kibble-Zurek theory [9, 10]. They predict a power law for the number, nex, of excitations: nex ∝ Nβν , (1) where β is the sweep rate through the critical point of N interacting qubits.

  2. For k = 0: nex ∝ β ν, as shown in Eq. p. 1

    Where the paper says it

    The case with k < 0 is also of interest because, for negative but small k, the system passes in the vicinity of the critical point, enhancing nonadiabatic effects. This 0+1-dimensional field theory is equivalent to onedimensional quantum mechanics, in which ϕ is identified with a coordinate x ≡ ϕ of a point particle having mass m and moving in the potential V (x, t), as shown in Fig.

    (7).

  3. For k > 0: nex ∝ n0 + cβν+, where n0 does not depend on β, unless one accounts for exponentially suppressed quantum over-barrier tunneling. p. 2

    Where the paper says it

    3 (ii) touching the critical boundary at k = 0; and (iii) the transition through the critical point at k > 0. We found that for k < 0 : nex ∝ β ν− e −∆/β , ∆ ∝ |k| µ− , (6) for k = 0 : nex ∝ β ν , (7) and for k > 0 : nex ∝ n0 + cβν+ , n0 ∝ k µ+ , (8) where c is some coefficient; n0 does not depend on β, unless we take into account exponentially suppressed quantum over-barrier tunneling, leading to extremely slow, logarithmic in β, decay of n0.

Numerical Validation and Critical Exponents

Numerical simulations of the LMGm confirm these predictions, with exponents derived as ν− = 1, ν = 3/4, ν+ ≈ 3/4, µ− = 2, and µ+ = 3/2. The analysis of the adiabatic invariant I(β) shows that for relatively fast transitions where ∆I >> I0, the exponent is ν+ = 0. p. 8

Where the paper says it

Summarizing, the power exponent ν+ emerges from two different effects that dominate depending on the relative values of ∆I and I0. For relatively fast transitions, when ∆I ≫ I0, we find ν+ = 0.

75. When I0 dominates, critical slowing down leads to an exponent of ν+ = 0. p. 8
Where the paper says it

On the other hand, when I0 dominates, the exponent becomes ν+ = 0. 8 because critical slowing down leads to noticeable changes in the potential well around the new global minimum before the system falls into it.

8.

Implications for Quantum Computation

The findings suggest that the performance of algorithms can be characterized by universality classes of phase transitions described by quantum field theories. The theory predicts that unwanted first-order critical points can be avoided by suppressing parameter uncertainties below a certain critical value in systems with well-mixed interactions. p. 8

Where the paper says it

The first-order transition can occur when unintentional energy barriers are introduced by small parameter uncertainties. Our theory predicts that unwanted first-order critical points can be avoided by suppressing these uncertainties below a certain critical value, at least in systems with well-mixed interactions.

This correspondence implies that the existence of new, faster algorithms can be inferred for some known computational problems based on these transition scalings. p. 8
Where the paper says it

Namely, field theories leading to different scalings of defects at quantum phase transitions should correspond, via the link to quantum annealing, to different convergence rates of computational algorithms. By developing this correspondence, we hope that the existence of new, faster algorithms can be inferred for some known computational problems.

Limitations and Future Directions

The theory suggests that integrability is a common feature of time-dependent field theories of quantum annealing, as the dynamics near the critical point are described by the integrable Painlevé-I equation. The paper notes that while the exponents 0. p. 4

Where the paper says it

We began with a state |ψ(−tmax)⟩ = |−S⟩, that is, “all spins pointing down” at an initial time of −tmax = −500/β. All results in this paper were calculated using S = 30, Bx = 1.

75 and 0.8 are close, a single power law with logarithmic correction provides a good fit to numerical data. p. 8
Where the paper says it

This is why a good fit to the numerical data for ∆I(β) can be obtained in Fig. 3 using a single power-law exponent supplemented by a logarithmic correction.

Further theoretical attention is needed regarding the emergence of integrability in these effective field theories. p. 2
Where the paper says it

(11) is that they describe well-mixed interactions. Another phenomenon that may need further theoretical attention is the emergence of integrability in the effective field theories of the passage through a critical point.

The paper also suggests that the presence of magnetic hysteresis supports the observation of first-order critical effects in D-Wave simulations. p. 8
Where the paper says it

Therefore, it is likely that, by attempting to detect the scaling of a second-order phase transition, these experiments observed first-order critical effects. Their presence in D-Wave simulations is also supported by observations of magnetic hysteresis [24].

Methods and Model Connection

The field Lagrangian in Eq. (4) is recovered from the quantum model when the number of spins is large and the coupling D is close to the critical value, i. p. 2

Where the paper says it

(4) is recovered from the quantum model in Eq. (3) when the number of spins is large and the coupling D is close to the critical value, D ≈ Bx, i.

e., D − Bx/Bx ≪ 1, and N >> 1. The Lagrangian parameters in Eqs. p. 2
Where the paper says it

, |D − Bx|/Bx ≪ 1, and N ≫ 1. The Lagrangian parameters in Eqs.

(4) and (5) are related to the parameters of the spin Hamiltonian via m = 1 / BxS, k = D − Bx / S, g = Bx / 2S3. p. 2
Where the paper says it

The Lagrangian parameters in Eqs. (4) and (5) are related to the parameters of the spin Hamiltonian (3) as m = 1 BxS , k = D − Bx S , g = Bx 2S3 .

p. 2
Where the paper says it

The Lagrangian parameters in Eqs. (4) and (5) are related to the parameters of the spin Hamiltonian (3) as m = 1 BxS , k = D − Bx S , g = Bx 2S3 .

The large N limit of LMGm simplifies the Hamiltonian to H = −βtSˆz − D2S−1 Sˆz2 − BxSˆx, where S is a large integer. p. 3
Where the paper says it

Thus, our following analytical predictions should be relevant and testable with accessible quantum annealing processors. Large N limit of LMGm The Hamiltonian of this model was defined in Eq.

The semiclassical approximation leads to a classical particle moving in a quartic potential Hcl(x, p, t) = p2 / 2m − kx2 / 2 + gx4 / 4 − βtx. p. 4
Where the paper says it

It is convenient to make a canonical transformation: φ → −p, n → x, which makes (x, p) a canonical pair of variables, such that x˙ = ∂pHcl, p˙ = −∂xHcl, where Hcl(x, p, t) = BxSp2 2 − βtx − D − Bx 2S x 2 + Bxx 4 8S3 . (17) Thus, the Hamiltonian acquires the form of a classical particle with mass m moving in a quartic potential: Hcl(x, p, t) = p 2 2m − kx2 2 + gx4 4 − βtx, (18) where parameters are related to the original quantum model via Eq.

Numerical Simulations Details

The numerical simulations were constructed using the Hamiltonian (13) as a sparse tridiagonal matrix and evolved the state ψ(t) using the matrix exponential. The initial state was set to ψ(−tmax)⟩ = -S⟩, meaning “all spins pointing down” at an initial time of −tmax = −500/β. p. 4

Where the paper says it

5 which can be efficiently done using the Crank-Nicolson update (given a small enough ∆t) and efficient C++ linear algebra solvers: I + i ∆t 2 H ψ(t + ∆t) = I − i ∆t 2 H ψ(t). We began with a state |ψ(−tmax)⟩ = |−S⟩, that is, “all spins pointing down” at an initial time of −tmax = −500/β.

The number of excitations is computed as nex = S − ⟨ψ(tmax)Sˆzψ(tmax)⟩. p. 4
Where the paper says it

0, and 3 × 106 time steps. Having completed the evolution, we compute the number of excitations as nex = S − ⟨ψ(tmax)|Sˆ z|ψ(tmax)⟩.

The final adiabatic invariant in the original variables is given by I(β) = ηθI(b(β)) = k3 / 2m1/2 g / βg1/2m1/2 k2. p. 5
Where the paper says it

Since H depends only on b, the final adiabatic invariant is also a function of only this combination: I = I(b). The adiabatic invariant in the original variables is then given by I(β) = ηθI(b(β)) = |k| 3/2m1/2 g I βg1/2m1/2 |k| 2 .

The calculation of the prefactor in Eq. p. 8
Where the paper says it

, the system spends an anomalously long time near the critical point. However, in our calculation of the coefficient I0 in Eqs.

(37) yields ⟨I⟩ = πb / 5√3 e − π√3 / 8b. p. 6
Where the paper says it

The leading exponent of the integral (taken as absolute square) is given by ⟨I⟩ ∝ e − 2 b Im R x0 0 (1+3x 2 m) 3/2 dxm, = e − 2 b R 1/ √ 3 0 (1−3s 2 ) 3/2 ds = e − π √ 3 8b . Calculating also the prefactor, we find ⟨I⟩ = πb 5 √ 3 e − π √ 3 8b .

Phase Transition Dynamics at k > 0

For the case where k > 0, the adiabatic limit is illustrated by stages (a-c) in Fig. 1(a-c). The crossing time tc is determined by setting the first derivative of V(x) to zero at x = xc, leading to tc = 2 / (3√3). p. 7

Where the paper says it

Hence, for these initial conditions, we should select the negative solution: xc = −1/ √ 3. The corresponding time moment, tc, of crossing the critical point is determined by setting the first derivative of V (x) to zero at x = xc: btc = −xc + x 3 c = 2 3 √ 3 .

The adiabatic invariant for this trajectory is calculated explicitly as I0 = 1 / (π Z x1 xc dx p(x)) = 2 / (33⁄2). p. 7
Where the paper says it

Given the analogy with classical particle motion, the kinetic energy is p 2/2, and p(x) = p 2[E − V (x)] = (x − xc) p (x − xc)(x1 − x)/2. (41) The adiabatic invariant for this trajectory can be calculated explicitly: I0 = 1 2π I p(x) dx = 1 π Z x1 xc dx p(x) = 2 3 3/2 .

The β-dependent contribution to the adiabatic invariant is given by ∆I = 0. p. 6
Where the paper says it

Having a time-dependent trajectory x(t), we calculated the final action as an integral of (x ′ ) 2 over time along a periodic trajectory. We then extracted the β-dependent part of the adiabatic invariant, ∆I, by fitting points at different β using Eq.

646 · β3⁄4m7/8 g5/8. p. 7
Where the paper says it

Therefore, the change in the adiabatic invariant can be found up to a numerical coefficient that was determined in [2]: ∆I = 0. 646 · β 3/4m7/8 g 5/8 .

This leads to the prediction that for relatively fast crossing rates, the scaling exponent is ν+ = ν = 3/4. p. 7
Where the paper says it

Analytical estimates produce slightly different predictions for the power exponent ν+, depending on the relative strengths of the two terms in Eq. First, consider the case of a relatively fast crossing rate, so that the constant contribution is negligibly small in comparison with purely dynamic effects.

When I0 dominates, the exponent becomes ν+ = 0. p. 8
Where the paper says it

75, and the β-dependent term arises from nonadiabatic effects analogous to those at second-order phase transitions. On the other hand, when I0 dominates, the exponent becomes ν+ = 0.

8 because critical slowing down leads to noticeable changes in the potential well around the new global minimum. p. 8
Where the paper says it

On the other hand, when I0 dominates, the exponent becomes ν+ = 0. 8 because critical slowing down leads to noticeable changes in the potential well around the new global minimum before the system falls into it.

Subcritical Regime Analysis

In the subcritical regime where k < 0, a single minimum exists, and the dynamics are approximated by a harmonic oscillator with frequency ω2(x) = 1 + 3x2m. The average change of the adiabatic invariant is given by an integral that can be calculated using the method of steepest descent, yielding ⟨I⟩ ∝ e − π√3 / 8b. p. 6

Where the paper says it

(30) and (32), this leads to ⟨I⟩ = 1 2 Z ∞ −∞ dxm r ω(xm) 2 e iθ(xm) 2 = 1 4 Z ∞ −∞ dxm p ω(xm)e i R xm −∞ ω(x) dt dx dx 2 = 1 4 Z ∞ −∞ dxm p ω(xm)e i b R xm −∞(1+3x 2 ) 3/2 dx 2 . Due to the smallness of b, this integral can be calculated by the method of steepest descent.

This leads to sub-critical exponents ν− and µ− listed in Eq. p. 6
Where the paper says it

Calculating also the prefactor, we find ⟨I⟩ = πb 5 √ 3 e − π √ 3 8b . (37) For b ∼ β/|k| 2 , we obtain the sub-critical exponents ν− and µ− listed in Eq.

(11). The final expression for the number of excitations is nex = 0. p. 0
Where the paper says it

Near the weakly first-order critical point, this level enters a region with a locally dense spectrum, where nonadiabatic excitations are highly probable. After leaving this critical region, the adiabaticity conditions are restored, and the system emerges on an adiabatic energy level carrying nex excitations above the final ground energy.

51 · Sβ / B3⁄2 x (Bx − D)1⁄2 e − 0. p. 6
Where the paper says it

(10) to find nex = 0. 51 · Sβ B 3/2 x (Bx − D) 1/2 e −0.

962·(Bx−D)2/β. p. 6
Where the paper says it

51 · Sβ B 3/2 x (Bx − D) 1/2 e −0. 962·(Bx−D) 2/β .

Conclusion and Universality

The power exponent ν+ emerges from two different effects that dominate depending on the relative values of ∆I and I0. This suggests that the essentially nonadiabatic effects near the critical point make a negligible contribution to ∆I in the slow-transition limit. p. 8

Where the paper says it

8 that we found numerically. Thus, we conclude that the essentially nonadiabatic effects near the critical point make a negligible contribution to ∆I in the slow-transition limit.

The paper concludes that field theories leading to different scalings of defects at quantum phase transitions should correspond to different convergence rates of computational algorithms. p. 8
Where the paper says it

Based on our findings, we speculate that the performance of algorithms more generally can be characterized by universality classes of phase transitions, which are described by quantum field theories. Namely, field theories leading to different scalings of defects at quantum phase transitions should correspond, via the link to quantum annealing, to different convergence rates of computational algorithms.

Model Parameters and Connections

The Lagrangian parameters in Eqs. (4) and (5) are related to the parameters of the spin Hamiltonian via m = 1 / BxS, k = D − Bx / S, g = Bx / 2S3. The value k = 0 separates the regimes with and without the first-order phase transition in LMGm, corresponding to the interaction strength D = Bx. p. 2

Where the paper says it

(4) and (5) are related to the parameters of the spin Hamiltonian (3) as m = 1 BxS , k = D − Bx S , g = Bx 2S3 . (10) The value k = 0 separates the regimes with and without the first-order phase transition.

The connection between the field Lagrangian and the quantum model is established when N is large and D ≈ Bx. p. 2
Where the paper says it

(9) In Methods, we show that the field Lagrangian in Eq. (4) is recovered from the quantum model in Eq.

Numerical Results Summary

Figure 3 shows results of numerical simulations of Newton’s equation x'' = bt + x − x3, which is equivalent to the evolution equations (19). The best fit to the power-law scaling for ∆I(β) suggests ν+ = 0. p. 6

Where the paper says it

A three-parameter fit with a logarithmic correction, ln(∆I) = ν+ ln β + a(ln ln(β))r , suggests that the true exponent is ν+ = 0. 8, but there is a logarithmic correction to this scaling, with a ≈ 0.

8, with a logarithmic correction. p. 6
Where the paper says it

However, the coefficient c showed an additional weak dependence on β. A three-parameter fit with a logarithmic correction, ln(∆I) = ν+ ln β + a(ln ln(β))r , suggests that the true exponent is ν+ = 0.

The numerical results confirm that the exponent ν+ is close to 0. p. 6
Where the paper says it

2 and r = 1. In any case, the numerical results confirm that ν+ is close to the value 0.

75, at which the second term in Eq. p. 6
Where the paper says it

In any case, the numerical results confirm that ν+ is close to the value 0. 75, at which the second term in Eq.

(27) does not depend on k. p. 2
Where the paper says it

3 (ii) touching the critical boundary at k = 0; and (iii) the transition through the critical point at k > 0. We found that for k < 0 : nex ∝ β ν− e −∆/β , ∆ ∝ |k| µ− , (6) for k = 0 : nex ∝ β ν , (7) and for k > 0 : nex ∝ n0 + cβν+ , n0 ∝ k µ+ , (8) where c is some coefficient; n0 does not depend on β, unless we take into account exponentially suppressed quantum over-barrier tunneling, leading to extremely slow, logarithmic in β, decay of n0.

The paper also notes that the subcritical passage of the model was studied previously by means of mean-field theory. p. 2
Where the paper says it

Two additional observations can be of theoretical interest. First, the subcritical passage of the model (3) was studied previously by means of mean-field theory, which considered the dynamics of a single spin in a regular field found self-consistently [16].

Acknowledgments and References

This work was carried out under the auspices of the U.S. DoE through the Los Alamos National Laboratory, operated by Triad National Security, LLC (Contract No. p. 8

Where the paper says it

ACKNOWLEDGMENTS This work was carried out under the auspices of the U. DoE through the Los Alamos National Laboratory, operated by Triad National Security, LLC (Contract No.

p. 8
Where the paper says it

ACKNOWLEDGMENTS This work was carried out under the auspices of the U. DoE through the Los Alamos National Laboratory, operated by Triad National Security, LLC (Contract No.

892333218NCA000001). p. 8
Where the paper says it

DoE through the Los Alamos National Laboratory, operated by Triad National Security, LLC (Contract No. 892333218NCA000001).

p. 8
Where the paper says it

DoE through the Los Alamos National Laboratory, operated by Triad National Security, LLC (Contract No. 892333218NCA000001).

The acknowledgments section lists numerous references supporting the theoretical framework and numerical studies. p. 8
Where the paper says it

We showed that the Suzuki–Zurek scaling can be inferred from quantum-coherent field theories, explaining its relevance to quantum annealing computations. Although quantum annealing deals with a combinatorially complex phase space, our approach is universal and straightforward for both analytical and numerical studies.

p. 8
Where the paper says it

We showed that the Suzuki–Zurek scaling can be inferred from quantum-coherent field theories, explaining its relevance to quantum annealing computations. Although quantum annealing deals with a combinatorially complex phase space, our approach is universal and straightforward for both analytical and numerical studies.

The extensive list of cited works demonstrates the broad context of this research in quantum annealing and phase transitions. p. 0
Where the paper says it

The driven Lipkin-Meshkov-Glick model (LMGm) serves as the minimal model of interacting Ising spins that demonstrates this behavior. INTRODUCTION Quantum annealing relies on the ability of a system to follow its instantaneous ground state while the Hamiltonian is slowly varied in time.

p. 0
Where the paper says it

The driven Lipkin-Meshkov-Glick model (LMGm) serves as the minimal model of interacting Ising spins that demonstrates this behavior. INTRODUCTION Quantum annealing relies on the ability of a system to follow its instantaneous ground state while the Hamiltonian is slowly varied in time.

The work was carried out under the auspices of the U. p. 8
Where the paper says it

By developing this correspondence, we hope that the existence of new, faster algorithms can be inferred for some known computational problems. ACKNOWLEDGMENTS This work was carried out under the auspices of the U.

S. DoE through the Los Alamos National Laboratory, operated by Triad National Security, LLC (Contract No. 892333218NCA000001). The acknowledgments section lists numerous references supporting the theoretical framework and numerical studies. The extensive list of cited works demonstrates the broad context of this research in quantum annealing and phase transitions.

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. Develop a predictive model for defect generation rates during quantum annealing by using field theory to identify power laws based on whether a first-order phase transition or second-order phase transition is occurring, as described in Eq (6), (7), and (8).

  2. Design an algorithm to detect the signature of first-order critical points in quantum annealing experiments by searching for sharp changes between predicted power exponents as continuous parameters are varied, which would be a signature of first-order critical points.

  3. Implement a method to avoid unwanted first-order critical points in computations by using the theory to determine the necessary suppression of parameter uncertainties below a certain critical value, based on the discussion in Section IV.

Abstract

Unlike second-order phase transitions, a first-order transition has a stage, in which a system is trapped in a metastable state. The decay of this state leads to abundant excitations over the ground state. We present a field theory for kinetics of defects emerging during quantum annealing computations. This theory predicts several power laws for the error generation rate during quantum annealing either through or near the first-order phase transition. Sharp changes between the power exponents are predicted for continuous parameter changes. Observation of this behavior would be a signature of first-order critical points and could help identify and avoid them for better computations. The driven Lipkin-Meshkov-Glick model (LMGm) serves as the minimal model of interacting Ising spins that demonstrates this behavior.

Sources

Related papers