Partial Reversibility and Counterdiabatic Driving in Nearly Integrable Systems
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: "Partial Reversibility and Counterdiabatic Driving in Nearly Integrable Systems".
Mira: Here's the gist of the paper. Partial Reversibility and Counterdiabatic Driving in Nearly Integrable Systems studies how reversible processes are possible in systems where integrability is broken by perturbations, exploring how approximate counterdiabatic driving can suppress dissipative losses Defining Adiabaticity and Reversibility Adiabatic (or reversible) processes are the key concept unifying our understanding of thermodynamics and dynamical systems <ref:2602.22317#pg2>.
Kai: In integrable dynamical systems, reversibility is understood as the conservation of the action variables <ref:2602.22317#pg3>.
Mira: Between these two idealized limits, however, things are much less clear <ref:2602.22317#pg4>.
Kai: The conventional definition of adiabaticity is reversibility – that is, if one executes some slow change of an external parameter of the system, the entropy should not increase <ref:2602.22317#pg4>.
Title and authors: Kai: So, we're looking at this paper called "Partial Reversibility and Counterdiabatic Driving in Nearly Integrable Systems". Mira, what's the big picture takeaway from the title itself?
Mira: It’s about how we can have reversible processes in systems where things aren't perfectly integrable anymore, and how some driving techniques can help fight the energy losses that usually happen when you mess with those systems.
Kai: Right. So, it tackles that gap between perfect reversibility in simple models and what happens when you introduce the kind of stuff we see in real experimental setups or complex quantum systems.
Lev: From my side, I'm always thinking about how these kinds of non-integrable perturbations show up in our actual error correction codes. If a process isn't perfectly reversible, it means some information is getting lost to the environment even when you try to keep things slow.
Kai: Exactly. So, the paper is setting up a test case by looking at two different types of nonlinear perturbations for these systems and seeing how they behave under different driving protocols.
Mira: That’s right. They set up three specific ramping protocols: Integrable to Integrable, Integrable to Nonintegrable, and Nonintegrable to Nonintegrable. These are ways you change the perturbation strength beta from a starting point beta i to an ending point beta f.
Lev: And they compare them based on the strength of that perturbation, using numbers like Protocol Model zero zero point two two nine for I-I, and then going up to Protocol Model five eight point eight five for N-N <ref:2602.22317#pg2>. That gives us a concrete way to test the limits of reversibility.
Kai: And what they’re trying to figure out is whether after these cycles, the final energy variance returns all the way back to zero, which would be the sign of perfect reversibility in that specific protocol.
Mira: They found that in one specific middle regime, between integrability being broken but not strong enough for complete chaos, like the I-N protocol, those fluctuations don't suppress down to a plateau. They attribute this failure to genuine irreversibility in those setups.
Lev: So they’re showing that there are specific conditions under which you just can't get rid of the dissipation by just changing how fast you drive the system.
The paper's summary: Kai: So, summarizing what they actually did in this paper, it looks like they started by testing toy models—two uncoupled harmonic oscillators with different nonlinear perturbations—to see if reversibility is even possible in these nearly integrable systems.
Mira: They explored the idea that adiabaticity means the entropy doesn't increase during a slow change of an external parameter, but they noted that when you move from regular to mixed phase space, it’s not immediately obvious if entropy can stay constant without some loss.
Lev: And they specifically looked at the I-I, I-N, and N-N protocols to see how the system behaves during these ramps in terms of energy fluctuations. The core finding they want us to hear is that for those intermediate I-N protocols, those fluctuations just don't settle down to a low level.
Kai: They concluded that this failure to suppress fluctuations points toward genuine irreversibility happening even when you’re trying to keep the driving slow enough, which they found happens in the case of the I-N protocol.
Mira: And then they brought in counterdiabatic driving as a way to try and fight those diabatic excitations. They used an approximate counterdiabatic Hamiltonian, HCD(t), defined as H(beta(t)) plus beta dot A beta of beta, where A is an adiabatic gauge potential.
Lev: The key here is that they use an expansion in Krylov space to approximate this gauge potential A, and when the driving is fast compared to the intrinsic time scale of the system, successive orders in that expansion can significantly suppress those fluctuations before they hit a plateau.
Kai: So, local counterdiabatic driving isn't a magic fix for everything; it works by suppressing some of those excitation noise from rapid driving, but there's a limit to how much it can help.
Mira: And that limit is important because they found that the fluctuations in the slow driving, which we call the large tau limit, actually provide a lower bound on how low those fluctuations can get.
Lev: That’s a crucial detail for hardware researchers because it means even if you use fast driving with counterdiabatic terms, you can't get better than what happens when you drive infinitely slowly; that slow limit sets the floor.
The paper's improvements: Kai: So, the authors suggest a few ways to think about this research and improve it. They highlight that partial reversibility in quantum many-body systems with large degeneracy is governed by nonperturbative effects beyond just the standard SW approximation during those I-N protocols.
Mira: That means if you have a perturbation strength epsilon f smaller than a critical value epsilon star, then the system can maintain reversibility even when levels hybridize, which is where things usually get messy.
Lev: But wait, what about real hardware? If we take this idea of maintaining reversibility during slow driving in degenerate systems and try to implement it, we need to look at how those blocks interact. The paper suggests that if you design the system so these symmetry blocks stay isolated during the slow change in parameter beta, then reversibility can be maintained because there are no transitions between those blocks.
Kai: That’s a way to engineer the physics to support reversibility instead of just trying to fight it with driving corrections. So, we design models where the specific symmetry sectors you care about remain separate while you ramp up your perturbation strength.
Mira: And they also discuss using a variational approach for local counterdiabatic driving when dealing with generic many-body systems, which gives us a practical protocol to control those irreversible losses that happen from rapid driving.
Lev: From an engineering standpoint, that variational approach would be something we could actually use to fight dissipation in real-world systems; it’s not just theoretical math anymore. But the paper also notes a limitation: local counteracting fast driving can only suppress fluctuations up to a certain plateau, beyond which local corrections don't help any further.
Kai: So, the core improvement they point out is that you need to look at nonperturbative effects when dealing with quantum systems with degeneracy and then engineer the system structure itself to avoid transitions between those sectors during slow driving.
Conclusion: Mira: To wrap up on this paper, the main conclusion is that breaking integrability in a system with degeneracies is fundamentally an irreversible process, driven by nonperturbative breakdown of approximate emergent symmetries during infinitely slow driving.
Kai: And they show that while local counterdiabatic driving can suppress some fluctuations from fast driving, those associated with the irreversibility they found provide a bound that you just can't beat.
Lev: So what this means for the hardware we build is that if you rely on these nearly integrable systems, you have to accept a certain level of unavoidable loss determined by that slow driving limit.
Mira: Exactly. But if your perturbation strength stays below epsilon star, and you engineer the system structure so those symmetry blocks don't hybridize during the ramp, then reversibility can be maintained under slow driving protocols.
Kai: It’s a nuanced picture—it’s not all about one solution. It shows that fighting dissipation requires understanding both the nonperturbative physics of the loss and engineering a specific structure to avoid certain transitions.
Lev: For us in error correction, it means we need to be very careful about how we design the Hamiltonian, making sure those blocks stay isolated when you're changing parameters slowly.
Mira: That’s the practical implication for us—designing models that respect those constraints gives us a path toward controlled evolution even in these complex settings.
Boston University
quant-ph, cond-mat.stat-mech, nlin.CD
Submitted: 2026-02-25
Updated: 2026-10-29
Comments: 11 pages, 6 figures
Journal ref: Phys. Rev. A 114, 042625 (2026)
DOI: 10.1103/4hkp-j534
Code: https://github.com/rohanb97/Classical_2D_
License: http://creativecommons.org/licenses/by/4.0/
Importance score: 92/100
The gist: Partial Reversibility and Counterdiabatic Driving in Nearly Integrable Systems studies how reversible processes are possible in systems where integrability is broken by perturbations…
Terminology
Summary
Partial Reversibility and Counterdiabatic Driving in Nearly Integrable Systems studies how reversible processes are possible in systems where integrability is broken by perturbations, exploring how approximate counterdiabatic driving can suppress dissipative losses
Defining Adiabaticity and Reversibility
Adiabatic (or reversible) processes are the key concept unifying our understanding of thermodynamics and dynamical systems <ref:2602.22317#pg2>. In integrable dynamical systems, reversibility is understood as the conservation of the action variables <ref:2602. p. 0 Partial Reversibility and Counterdiabatic Driving in Nearly Integrable Systems
Rohan Banerjee,∗ Shahyad Khamnei, Anatoli Polkovnikov, and Stewart Morawetz†
Department of Physics, Boston University, Boston, Massachusetts 02215, USA
(Dated: July 1, 2026)
Adiabatic (or reversible) processes are the key concept unifying our understanding of thermodynamics and dynamical systems. Reversibility in the thermodynamic sense is understood as entropypreserving processes, such as in the idealized Carnot engine, whereas in integrable dynamical systems
it is understood as the conservation of the action variables. Reversibility in the thermodynamic sense is understood as entropypreserving processes, such as in the idealized Carnot engine, whereas in integrable dynamical systems
it is understood as the conservation of the action variables. Between these two idealized limits, however, things are much less clear. If integrability is broken, but not strongly
enough to induce complete ergodicity, there is no universally agreed upon definition of adiabaticity. The conventional definition of adiabaticity is reversibility [6] – that
is, if one executes some slow change of an external parameter of the system, the entropy should not increase. The conventional definition of adiabaticity is reversibility [6] – that
is, if one executes some slow change of an external parameter of the system, the entropy should not increase. When this change is accompanied by a change in the
phase space from regular to mixed (i. When this change is accompanied by a change in the
phase space from regular to mixed (i. when breaking
integrability), it is not clear a priori if such a process is
possible without an increase of entropy.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Testing Reversibility with Toy Models
The authors examine two uncoupled harmonic oscillators subjected to different nonlinear perturbations, one integrable and another nonintegrable for nonzero perturbation strength β <ref:2602.22317#pg6>. They define three separate protocols based on the integrability of the system—Integrable to Integrable (I-I), Integrable to Nonintegrable (I-N), and Nonintegrable to Nonintegrable (N-N) <ref:2602. p. 1 We therefore define an integrable to
nonintegrable (I-N) protocol where driving weakly breaks
integrability, and a nonintegrable to nonintegrable (N-N)
protocol which is within the strongly integrability breaking regime where we expect thermodynamic reversibility
to apply. To characterize the behaviour of the three systems in
the slow-driving limit, we simulate the evolution of an
initially microcanonical ensemble with fixed energy E0 =
1 under driving of an external parameter β(t). In the inset, the fluctuations are shown immediately after the initial ramp to βf , with estimates given by
SW approximation shown as dashed lines. Reversible evolution implies that after any cyclic protocol, the state returns back to itself such that the final
energy variance must be zero. In between these two regimes the situation is more
complicated. If integrability is broken, but not strongly
enough to induce complete ergodicity, there is no universally agreed upon definition of adiabaticity.Where the paper says it
Where the paper says it
Where the paper says it
Efficacy of Counterdiabatic Driving
The focus shifts to suppressing diabatic excitations through approximate counterdiabatic (CD) driving, defined by the Hamiltonian HCD(t) = H(β(t)) + β˙Aβ(β(t)) <ref:2602.22317#pg3>. Approximations for the adiabatic gauge potential Aβ are necessary, and they use an expansion in Krylov space to approximate A (l)β <ref:2602. p. 2 However, as the exact Aβ does not even exist for a
generic classically chaotic system [51], approximations
are necessary. Following previous work [17, 22, 52], we
use an expansion in Krylov space. When the driving is fast compared to the intrinsic time
scale of the system (i. τ ≫ 1/ω), successive orders
in the Krylov space expansion can significantly suppress
fluctuations, before reaching a plateau.Where the paper says it
Where the paper says it
Partial Reversibility in Quantum Systems
The analysis extends to quantum many-body systems, considering a Hamiltonian HˆQ that possesses degenerate symmetry sectors <ref:2602.22317#pg2>. The authors argue that the physics behind partial reversibility is much the same, where nonperturbative effects beyond the SW approximation are responsible for any irreversibility during the I–N protocol <ref:2602. p. 3 els in different blocks start to cross. We will argue that the physics behind partial reversibility is much the same, where non-perturbative effects beyond the SW approximation are responsible for
any irreversibility during the I–N protocol. 2
a nonlinear perturbation, one which is integrable and
another which is nonintegrable for nonzero perturbation
strength β. HI =
p
2
x + p
2
y
2
+
x
2 + y
2
2
+ β
(x
2 + y
2
)
2
4
(1)
HNI =
p
2
x + p
2
y
2
+
x
2 + y
2
2
+ β
x
2y
2
2
(2)
In each case, the driving protocol consists of ramping β
from an initial value βi to a final value βf . Within the
SW approximation, as the strength of the perturbation ϵ
is increased from 0, the initially degenerate levels begin to
split. There will be some value ϵ
∗
(see End Matter) where
the split levels in different sectors begin to hybridize, and
the SW approximation breaks down. ϵ
∗ < ϵf ≪ h, global
thermalization between blocks sets in. At this point
the dynamics becomes irreversible and the cyclic ramp
leads to a non-zero energy variance even in the adiabatic
limit.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Conclusion on Control Protocols
The study concludes that breaking integrability in a system with degeneracies is a fundamentally irreversible process <ref:2602.22317#pg2>. The underlying reason is that such models can be characterized by approximate emergent symmetries, whose nonperturbative breakdown leads to an associated increase in entropy even in the infinitely slow driving limit <ref:2602. p. 3 We show
that breaking integrability in a system with degeneracies is a fundamentally irreversible process, and provide
numerical evidence supporting the claim for both classical and quantum toy models. The underlying reason
is that such models can be characterized by approximate
emergent symmetries, whose nonperturbative breakdown
leads to an associated increase in entropy even in the infinitely slow driving limit. The underlying reason
is that such models can be characterized by approximate
emergent symmetries, whose nonperturbative breakdown
leads to an associated increase in entropy even in the infinitely slow driving limit. We then test the efficacy of
local counterdiabatic driving in the toy model in different regimes, and find that while it can suppress some
fluctuations from fast driving, those associated with irreversibility provide a bound that cannot be beaten.Where the paper says it
Where the paper says it
End Matter Details
The Schrieffer-Wolff transformation is used to obtain an effective Hamiltonian HSW(β) which approximates H0 + βV, but where H0 is an exactly conserved quantity <ref:2602.22317#pg2>. The energy fluctuations derived within the Schrieffer-Wolff approximation are compared with numerical results, showing excellent agreement at smaller βf <ref:2602. p. 6 The Schrieffer-Wolff analytical predictions (solid
lines), compared with numerical results (dots) for the energy
fluctuations from degeneracy-lifting perturbation in the integrable and nonintegrable models. The agreement is excellent
at smaller βf . (6) is
driven to increasingly strong perturbations, and ϵf increases beyond ϵ
∗
, the dynamics become irreversible and
result in some non-zero σ
∗
E in the slow-driving limit. For
HˆQ, we can roughly calculate ϵ
∗ as the point at which
the typical fluctuations (roughly the standard deviation)
of the energy spectrum within a block is equal to the
spacing between blocks, i.Where the paper says it
Where the paper says it
Supplemental Material Notes
A randomized waiting time in the middle of the cyclic protocol results in smooth (exponential) decay in fluctuations in the slow driving limit, as expected for an integrable system <ref:2602.22317#pg2>. The detailed derivation of CD driving involves minimizing an action Sβ(Aβ) =∥Gβ∥2 + µ2∥Aβ∥2; Gβ = ∂βH − 1/iħ[Aβ, H] <ref:2602. p. 8 Instead, as was shown in Ref. [S16], solving this equation is equivalent to variationally
minimizing an action
Sβ(Aβ) = ∥Gβ∥
2 + µ
2
∥Aβ∥
2
; Gβ = ∂βH −
1
iℏ
[Aβ, H] (S3)
where we have defined the norm
∥O∥
2 = ⟨O
2
⟩ − ⟨O⟩
2
(S4)
with ⟨O⟩ = Tr(ρO), and µ is some small number added for regularization in case there are many degenerate solutions. The Chebyshev polynomials are defined given the recursive relation:
T0(x) = 1
T1(x) = x
Tn+1(x) = 2xTn − Tn−1
To formulate something in terms of operators, we can define a Super-Louvillian Operator L = {H, ·}, which corresponds to the multiplication of the phase space function by ω in frequency space. Recursive application of this defines
a set of operators which are orthogonal with respect to a particular measure:
Q0 = ∂βH
Q1 = L{Q0}
Qn+1 = 2L{Qn} − Qn−1
Using this modified basis, we can construct the AGP as follows:
A
(ℓ)
β =
X
ℓ
k=1
γkQ2k−1 (S7)
where ℓ is the order at which we truncate the expansion, and γk are the variational parameters that are computed
from minimizing the action in Eq. Higher Order CD Protocols
In the main text, we show up to five orders in the Krylov space expansion for local CD driving. To show that the
performance does indeed plateau, we show the energy variance resulting from up to seven orders in the Krylov space
expansion in Fig.Where the paper says it
Where the paper says it
Where the paper says it
References
[1] M. A Nielsen and I. p. 4 edu
† morawetz@bu. Nielsen and I. Nielsen and I. Chuang, Quantum Computation
and Quantum Information, 10th ed. Chuang, Quantum Computation
and Quantum Information, 10th ed. (Cambridge university press, Cambridge, 2010).Where the paper says it
Where the paper says it
Where the paper says it
[2] T. Albash and D. p. 4 (Cambridge university press, Cambridge, 2010). Albash and D. Albash and D. Lidar, Reviews of Modern Physics
90, 015002 (2018).Where the paper says it
Where the paper says it
[3] C. Wittig, The Journal of Physical Chemistry B 109, 8428 (2005), https://doi. p. 4 Lidar, Reviews of Modern Physics
90, 015002 (2018). Wittig, The Journal of Physical Chemistry B 109,
8428 (2005), https://doi. Wittig, The Journal of Physical Chemistry B 109,
8428 (2005), https://doi. 1021/jp040627u.Where the paper says it
Where the paper says it
[4] V. I. Arnold, Mathematical Methods of Classical Mechanics (Springer New York, 1989) <ref:2602. p. 4 1021/jp040627u. Arnold, Mathematical Methods of Classical Mechanics (Springer New York, 1989).Where the paper says it
[5] P. Lochak, Multiphase Averaging for Classical Systems: With Applications to Adiabatic Theorems, Applied Mathematical Sciences Ser No. p. 4 Arnold, Mathematical Methods of Classical Mechanics (Springer New York, 1989). Lochak, Multiphase Averaging for Classical Systems:
With Applications to Adiabatic Theorems, Applied Mathematical Sciences Ser No. Lochak, Multiphase Averaging for Classical Systems:
With Applications to Adiabatic Theorems, Applied Mathematical Sciences Ser No. 72 (Springer New York, New
York, NY, 1988).Where the paper says it
Where the paper says it
[6] L. Landau and E. p. 0 The great disadvantage to which adiabatic processes
are subject is the requirement of long wait times. This
condition will be most familiar to many readers in the
quantum context, such as the famous Landau-Zener
problem [3], where in order to remain adiabatic the driving rate should be smaller than the minimum value set by
the energy gap (squared). Landau and E. Lifshitz, Statistical Physics: Volume 5,
v. Lifshitz, Statistical Physics: Volume 5,
v. 5 (Butterworth-Heinemann, 2013).Where the paper says it
Where the paper says it
Where the paper says it
[7] D. Guéry-Odelin, A. p. 4 5 (Butterworth-Heinemann, 2013). Guéry-Odelin, A. Guéry-Odelin, A. Ruschhaupt, A. Ruschhaupt, A. Torrontegui, S. Torrontegui, S. Martínez-Garaot, and J. Martínez-Garaot, and J. 91, 045001 (2019).Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
[8] T. Hatomura, Journal of Physics B: Atomic, Molecular and Optical Physics 57, 102001 (2024) <ref:2602. p. 4 91, 045001 (2019). Hatomura, Journal of Physics B: Atomic, Molecular
and Optical Physics 57, 102001 (2024).Where the paper says it
[9] C. W. Duncan, P. M. Poggi, M. Bukov, N. T. Zinner, and S. p. 4 Hatomura, Journal of Physics B: Atomic, Molecular
and Optical Physics 57, 102001 (2024). Zinner,
and S. Zinner,
and S. Campbell, PRX Quantum 6, 040201 (2025).Where the paper says it
Where the paper says it
[10] A del Campo, Physical Review Letters 111, 100502 (2013) <ref:2602.22317#pg5>
[13] E. Boyers, M. Pandey, D. K. Campbell, A. p. 4 Zinner,
and S. Campbell, PRX Quantum 6, 040201 (2025). Partial Reversibility and Counterdiabatic Driving in Nearly Integrable Systems
Rohan Banerjee,∗ Shahyad Khamnei, Anatoli Polkovnikov, and Stewart Morawetz†
Department of Physics, Boston University, Boston, Massachusetts 02215, USA
(Dated: July 1, 2026)
Adiabatic (or reversible) processes are the key concept unifying our understanding of thermodynamics and dynamical systems. Reversibility in the thermodynamic sense is understood as entropypreserving processes, such as in the idealized Carnot engine, whereas in integrable dynamical systems
it is understood as the conservation of the action variables. Polkovnikov,
D. Sels, and A. Sels, and A. Sushkov, Physical Review A 100,
012341 (2019).Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
[16] D Sels and A Polkovnikov, Proceedings of the National Academy of Sciences 114, 10.1073/pnas.1619826114 (2017) <ref:2602. p. 4 Polkovnikov, Proceedings of the National
Academy of Sciences 114, 10. 1619826114
(2017).Where the paper says it
[29] S Morawetz and A Polkovnikov, PRX Quantum 6, 040320 (2025) <ref:2602.22317#pg5>
[33] C Jarzynski, Physical Review A 88, 040101 (2013) <ref:2602.22317#pg5>
[37] A Zhong, B. Kuznets-Speck, and M. p. 4 Jarzynski, Physical Review
Letters 100, 190601 (2008). Kuznets-Speck, and M. Kuznets-Speck, and M. DeWeese, Physical Review E 110, 034121 (2024).Where the paper says it
Where the paper says it
[49] J R Schrieffer and P A Wolff, Physical Review 149, 491 (1966) <ref:2602.22317#pg4>
[50] S Bravyi, D P DiVincenzo, and D Loss, Annals of Physics 326, 2793 (2011) <ref:2602.22317#pg5>
[51] C Jarzynski, Physical Review Letters 74, 1732 (1995) <ref:2602.22317#pg5>
[52] B Bhattacharjee, A Lanczos approach to the Adiabatic Gauge Potential (2023), arXiv:2302.07228 <ref:2602. p. 0 As a testing ground, we define several illustrative
toy models whose behaviour in this limit is qualitatively
different. We consider two uncoupled harmonic oscillators with
1
arXiv:2602.Where the paper says it
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
-
Counterdiabatic driving can suppress diabatic excitations resulting from external parameter changes at a finite rate, leading to perfectly adiabatic evolution regardless of how fast parameter change is made. This enables AI systems to evolve along desired trajectories in phase space without incurring uncontrolled losses during non-adiabatic transitions.
-
Local counterdiabatic driving can suppress fluctuations associated with irreversible losses, but it cannot beat the bound set by the slow-driving limit result, which is consistent with the observation that
the fluctuations in the slow driving (large tau) limit provide a lower bound.
This allows for more precise control of system evolution when fast driving is employed. -
For quantum many-body systems with large degeneracy, partial reversibility suggests that
nonperturbative effects beyond the SW approximation are responsible for any irreversibility during the I–N protocol,
implying that AI models starting in degenerate symmetry sectors can maintain reversibility under slow driving if perturbation strength is below a critical value epsilon star. -
AI systems modeled by Hamiltonians with degenerate symmetry blocks can be engineered such that they exhibit reversible evolution when driven slowly, as the absence of transitions between blocks leads to
reversibility
becausethe absence of transitions between blocks leads to reversibility.
This suggests designing models where specific symmetry sectors remain isolated during slow parameter changes. -
By using a variational approach for local counterdiabatic driving, AI systems can be equipped with an approximate protocol that provides control over irreversible losses in generic many-body systems. This allows the system to
fight dissipative losses resulting from rapidly driving these kinds of systems.
Abstract
Adiabatic (or reversible) processes are the key concept unifying our understanding of thermodynamics and dynamical systems. Reversibility in the thermodynamic sense is understood as entropy-preserving processes, such as in the idealized Carnot engine, whereas in integrable dynamical systems it is understood as the conservation of the action variables. Between these two idealized limits, however, things are much less clear. In this work, we study several toy models to assess the extent to which reversible processes are even possible in this regime. We then explore how the dissipative losses resulting from rapidly driving these kinds of systems can be fought by approximate counterdiabatic driving. Finally, we provide numerical evidence that much of the phenomenology should be the same for quantum many-body systems with large degeneracy in the presence of integrability breaking perturbations.
Sources
- Floquet-engineering counterdiabatic protocols in quantum many-body systems
- Shortcuts to Analog Preparation of Non-Equilibrium Quantum Lakes
- Fighting Exponentially Small Gaps by Counterdiabatic Driving
- Digitized counterdiabatic quantum critical dynamics
- Counterdiabatic Hamiltonian Monte Carlo
- Geometry and non-adiabatic response in quantum and classical systems
- A Lanczos approach to the Adiabatic Gauge Potential
- Defining classical and quantum chaos through adiabatic transformations
Related papers
- Computational regimes in matrix-product-state-based quantum trajectory simulations
- Toward chemically accurate quantum simulations on quantum hardware: A quantum-HPC hybrid approach
- Transversal surface-code game with reconfigurable qubits
- Fulqrum: a generalized framework for quantum subspace diagonalization
- Flux-Activated Resonant Control of a Bosonic Quantum Memory
- Machine learning the effects of many quantum measurements