A Counterexample to the Weak Gibbs and Regularity Conjectures
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Mathematical Physics. The day in mathematical physics: rigorous models of physical systems.
Kai: We're looking at the paper: "A Counterexample to the Weak Gibbs and Regularity Conjectures".
Mira: Here's the gist of the paper. A counterexample to the weak Gibbs and regularity conjectures is provided by a spin-one system on a square lattice with nearest-neighbor interaction and a uniform on-site field, demonstrating that its unique KMS state is nonregular and not weak Gibbs The Nonregularity Result The paper constructs a translation-invariant nearest-neighbor quantum spin system whose unique KMS state is nonregular The authors establish for a KMS state ω and an explicit translation-invariant state ν the strict inequality that disproves the regularity conjecture This is achieved by showing that lim sup k→∞ D(νΛLk ωΛLk) ΛLk ≤ pH(β) + 2βh - δ < pH(β) + 2βh Model Definition and Properties The model is defined on the square lattice Z squared, with a local Hilbert space C cubed and a Hamiltonian HZ involving hopping terms Fxy that turn one particle into two along the bond and two into one in the opposite direction The Hamiltonian conserves the dyadically weighted particle number CZ, which is invariant under HZ The existence of a unique KMS state ω for small inverse temperature β and arbitrary field strengths h is established Proof Strategy via Gibbs Sampler The proof proceeds by first proving uniqueness of the KMS state, then testing regularity against the empty product state ν The relative entropy density is defined as s(ν ω):= lim Z↗Zd D(νZ ωZ) Z = lim Z↗Zd trZ (log νZ − log ωZ) The authors use a Gibbs sampler to establish uniqueness and then bound the relative entropy between ω and ν using an imaginary-time return amplitude Key Technical Steps in the Bound The comparison between the empty-region probability and the return amplitude yields a bound on D(νΛL ωΛL) involving log ZΛL − log GΛ+L + O(L) By choosing a sufficiently large field h, the authors show that-log GΛ+Lk ≤ (2βh - δ)Λ+Lk This leads to the final contradiction where lim sup k→∞ D(νΛLk ωΛLk) ΛLk ≤ pH(β) + 2βh - δ < pH(β) + 2βh Invariance of the KMS State The authors prove that every translation-invariant state is KMS and that the unique KMS state ω is invariant under each limiting update Φx The invariance of the physical evolution implies that for every β-KMS state φ, φ(A) = φ(Tt(A)) → ω(A) as t → ∞ This demonstrates that omega is the unique KMS state Since the KMS state in the example is nonregular, it is also not weak Gibbs This result builds upon previous work that established failure of regularity in models with a unique KMS state The paper concludes that ω is the unique KMS state and every translate of ω is again KMS References
Title and authors: Kai: So, this paper is titled "A Counterexample to the Weak Gibbs and Regularity Conjectures," and it's about a spin-one system on a square lattice with nearest-neighbor interaction and a uniform on-site field. We're looking at how unique states in these systems behave when you try to extend them to infinite space, which is really important for understanding statistical mechanics in many quantum contexts.
Mira: Exactly, and the authors are showing that this specific system has a unique KMS state that doesn't meet the criteria for being regular or even weak Gibbsian. It’s about breaking those conjectures by finding a concrete example where things don't behave as expected when you take the thermodynamic limit.
Lev: From my side, I see this as a test case for how robust these conjectures are when you move beyond simple models, and this specific spin-one system is chosen because it exhibits some interesting dynamics with those hopping terms that turn one particle into two along the bond and two into one in the opposite direction.
Kai: So, the paper then goes on to summarize what they did. They're constructing this translation-invariant nearest-neighbor quantum spin system and establishing a strict inequality between the relative entropy density of a state called nu and our unique KMS state omega, which disproves the regularity conjecture.
Mira: That inequality is key because it shows that for some sequence of boxes with side length L approaching infinity, the difference in information between these two states doesn't go to zero fast enough compared to what you'd expect from a regular state. It’s this strict inequality: lim sup k to infinity D(nu lambda l k omega lambda l k) abs lambda l k <= p h of beta + two beta h minus delta < p h of beta + two beta h <ref:2611.06617#pg1>.
Lev: That bound is what really matters for me, because it gives us a way to quantify the failure. It shows that the relative entropy density doesn't stay bounded by the free energy excess plus some boundary term as you go to infinity, which is exactly what regular states are supposed to satisfy.
Kai: So, on top of this disproving regularity, they also look at weak Gibbsianity. They show that this unique KMS state omega is not weak Gibbs either because there isn't a sequence of constants c z greater than or equal to zero for finite lattice subsets z such that the relative entropy density is bounded by those exponentials in a way that the limit of c z divided by the size of the subset goes to zero.
Title and authors: Mira: Which means they are showing that this state doesn't have that kind of local control over its probability distribution as you look at larger and larger regions, which is what weak Gibbsianity requires. This is a stronger property than regularity, so disproving it means we've found a state that fails both conditions.
Lev: For practical applications on real quantum hardware, this means if you were to try and simulate this system using some form of Gibbs sampling or Monte Carlo method, the convergence properties wouldn't be as nice as we hope for a regular system. You might run into issues with the error scaling that these bounds suggest.
Kai: They set up their proof strategy by first proving uniqueness of the KMS state and then testing its regularity against the empty product state nu. They use a Gibbs sampler to establish that uniqueness, which is a solid starting point for this kind of analysis.
Mira: The relative entropy density they define is essentially the limit of the difference in logarithms between these states as you take larger and larger volumes. It’s how they measure the divergence between omega and nu when you scale up, but it's hard to control that term without regularity.
Lev: And then they use an imaginary-time return amplitude, which I think is where they get their main technical handle on the comparison error term. That return amplitude relates the probability of returning to a region with itself in imaginary time to the partition function terms, and that's what lets them build that specific bound.
Kai: The key technical step involves comparing this return amplitude with a simpler expression involving log of the partition function and this return amplitude itself, which leads to an error term bounded by the size of the box plus some small constant delta.
Mira: By choosing a sufficiently large on-site field strength h, they manage to show that that part involving the imaginary time return amplitude is actually smaller than what they need it to be for a regular state, which yields that final contradiction with the strict inequality.
Lev: The paper’s limitation is that this construction relies on finding a specific spin-one system and tuning the field strength h high enough to make that bound work, so it doesn't immediately tell us if this happens for *all* translation-invariant systems. It’s a very specific counterexample.
Title and authors: Kai: But what it does mean for us is that we have a concrete mathematical object, this spin-one system, where the unique KMS state is demonstrably nonregular and not weak Gibbs. It moves the discussion from theoretical possibility to a specific physical realization that we can analyze.
Mira: The implication is that whatever mechanisms cause this failure in this model—the hopping terms and the on-site field—those mechanisms are powerful enough to prevent states from being regular, which suggests these conjectures might not hold universally for all translation-invariant systems.
Lev: For error correction research, it’s a warning that if you're designing protocols based on these states, you need to be very careful about the assumptions about the underlying state structure because the error scaling will be worse than what a regular model predicts.
Kai: So, to wrap up this part of the discussion, we have established this spin-one system as a counterexample to both weak Gibbsianity and regularity conjectures by showing the relative entropy density fails to satisfy those expected bounds in the thermodynamic limit.
Mira: And they did it by constructing an explicit inequality using a Gibbs sampler and bounding the imaginary time return amplitude against a specific form involving h, which leads directly to that strict inequality we discussed earlier.
Lev: It really highlights how delicate these conjectures are, because even small perturbations, like the on-site field strength h in this case, can be enough to push a system out of the regular class.
Kai: So this paper is a precise demonstration that nonregularity is possible for translation-invariant states in these quantum systems and it forces us to re-evaluate our assumptions about what we expect from unique KMS states.
Mira: And looking ahead, they suggest improvements by establishing the relationship between the relative entropy density and the free energy excess plus pressure contribution, which provides a clearer way to compare any given state against a regular one.
Lev: That comparison is useful for error correction because if you can bound this difference tightly, you can bound the deviation of your actual state from an idealized, regular one.
Kai: So to conclude this segment on the paper itself, we've seen how they used a specific spin-one model to construct a state that fails both weak Gibbsianity and regularity by using a tight bound on the return amplitude against a specific term involving the field strength h.
Mira: And as we move toward the next part, we’re going to look at what they suggest as improvements for this work and what that means for future research in this area.
Lev: We'll talk about those technical suggestions and how they might translate into something more usable for actual quantum experiments.
The paper's summary: Kai: So, we're looking at this paper that builds on earlier work about unique KMS states in quantum systems and shows that this specific spin one system has a unique KMS state that isn't regular or weak Gibbsian.
Mira: Basically, they construct a translation invariant nearest neighbor quantum spin system and prove there is a strict inequality between the relative entropy density of their unique KMS state omega and some other state nu.
Lev: That inequality is what really disproves the regularity conjecture by showing that for some sequence of boxes with side length L going to infinity, the difference in information between these two states doesn't go to zero fast enough compared to what a regular state would expect.
Kai: It's this bound they establish: the limit superior of the relative entropy density times the size of the box is less than or equal to some value involving p h of beta plus two beta h minus delta, but strictly less than p h of beta plus two beta h.
Mira: That strict inequality is what really matters because it shows that for this state, you can't find a constant that controls the relative entropy density in the way weak Gibbsianity requires.
Lev: And it’s not just about regularity, either; they also show this unique KMS state omega is not weak Gibbs because there isn't a sequence of constants c z greater than zero for finite lattice subsets z such that the relative entropy density is bounded by those exponentials in the required way.
Kai: So, to summarize, they found a concrete example where the unique equilibrium state fails both regularity and weak Gibbsianity by setting up this specific comparison using a Gibbs sampler and bounding that imaginary time return amplitude against a term involving the field strength h.
Mira: The implication is that whatever mechanisms cause this failure in the model—the hopping terms and the on-site field—those mechanisms are powerful enough to prevent states from being regular, which suggests these conjectures might not hold universally for all translation invariant systems.
Lev: For us, as error correction researchers, this means if you're designing protocols based on these states, you need to be very careful about the assumptions about the state structure because the error scaling will be worse than what a regular model predicts.
Kai: It’s a warning that nonregularity is possible for translation invariant states in these quantum systems and it forces us to re-evaluate our assumptions about what we expect from unique KMS states.
Mira: Looking ahead, they suggest improvements by establishing the relationship between the relative entropy density and the free energy excess plus pressure contribution, which provides a clearer way to compare any given state against a regular one.
Lev: That comparison is useful for error correction because if you can bound this difference tightly, you can bound the deviation of your actual state from an idealized, regular one.
Kai: So we've established this spin one system as a counterexample to both weak Gibbsianity and regularity by showing the relative entropy density fails to satisfy those expected bounds in the thermodynamic limit.
Mira: And they did it by constructing an explicit inequality using a Gibbs sampler and bounding that imaginary time return amplitude against a specific form involving h, which leads directly to that strict inequality we discussed earlier.
Lev: It really highlights how delicate these conjectures are because even small perturbations, like the on-site field strength h in this case, can be enough to push a system out of the regular class.
The paper's improvements: Tom: So we're looking at what the authors suggest as improvements for their work on this spin one system that breaks the regularity and weak Gibbsianity conjectures.
Kai: They propose establishing a clearer relationship between that relative entropy density they calculated and something called the free energy excess plus pressure contribution.
Mira: That's key because it gives a better tool to compare any given state against what we expect from a regular state, rather than just comparing it to one specific reference state like nu.
Lev: It’s about making that comparison more direct, so if you want to know how far your actual physical system is from being regular, this new relationship helps you bound that deviation more tightly.
Kai: Essentially, they are trying to turn the abstract entropy calculation into a way of measuring the distance from regularity using thermodynamic quantities we can actually compute.
Mira: It’s about moving beyond just showing a state isn't regular and giving us a quantitative measure of *how* far it is from being regular in terms of those measurable quantities.
Lev: For error correction, that kind of tighter bound is what you need because you need to know the precise overhead when trying to simulate or approximate these states on real hardware.
Kai: It sounds like they're giving us a more practical way to use this mathematical result in building better quantum simulations or understanding physical systems that aren't perfectly ordered.
Mira: And that leads into another area where they suggest looking at how the structure of those patterns, maybe related to the five vertex model we saw on the other papers, might influence these state properties.
Lev: They hint at connecting this spin one system to some of those more complex lattice models, which would be interesting because those other models often have much harder physics.
Kai: So they're not just stopping at proving failure; they're suggesting the next step is to see if we can link this nonregularity to other known challenging quantum systems.
Conclusion: Kai: So we've seen how this paper, "A Counterexample to the Weak Gibbs and Regularity Conjectures," uses a specific spin one system to show that its unique KMS state fails both regularity and weak Gibbsianity.
Mira: It’s a concrete demonstration that you can have a unique equilibrium state in a quantum lattice model that still doesn't behave according to the expected rules for regular or weak Gibbs states when you take the thermodynamic limit.
Lev: From my side, it means if we're building error correction codes based on these states, we have to be extremely careful because the way errors scale is going to be worse than what a regular model predicts.
Kai: It really forces us to rethink our assumptions about what properties a unique KMS state must possess when it exists in infinite space.
Mira: The authors did this by constructing that specific inequality using the Gibbs sampler and bounding that imaginary time return amplitude against a term involving the field strength h.
Lev: That reliance on tuning the field strength h to get that bound is something we need to keep in mind for any simulation, because finding a suitable h can be tricky in practice.
Kai: So, this paper gives us a concrete object where nonregularity is possible and it shows the failure point clearly with that strict inequality.
Mira: It’s not just about one model; it points toward deeper structural issues in how these quantum lattice systems handle translation invariance when you push them to the limit.
Lev: I wonder if this specific type of system, with its hopping terms and on-site field, is representative of more general classes of models that we might encounter in real experiments.
Kai: Exactly; it’s a warning that nonregularity is possible for translation invariant states in these quantum systems and it forces us to re-evaluate our assumptions about what we expect from unique KMS states.
Mira: Moving forward, the authors are suggesting they establish a clearer way to compare that entropy density against free energy excess plus pressure contribution, which is a better tool than just comparing it to another state like nu.
Lev: That comparison is useful for error correction because if you can bound that difference tightly, you can bound the deviation of your actual state from an idealized, regular one.
Kai: It sounds like they're giving us a more practical way to use this mathematical result in building better quantum simulations or understanding physical systems that aren't perfectly ordered.
Mira: They are trying to move beyond just showing a state isn't regular and giving us a quantitative measure of *how* far it is from being regular in terms of those measurable quantities.
Lev: And they hint at connecting this spin one system to some of those more complex lattice models, maybe related to those other papers we've looked at, which would be interesting because those other models often have much harder physics.
Marius Wesle
University of Tübingen
math-ph, math.MP, quant-ph
Submitted: 2026-11-04
Updated: 2026-11-04
Comments: 19 pages, 2 figures
License: http://creativecommons.org/licenses/by/4.0/
Importance score: 62/100
The gist: A counterexample to the weak Gibbs and regularity conjectures is provided by a spin-one system on a square lattice with nearest-neighbor interaction and a uniform on-site field…
Key concepts
- KMS State
- A KMS (Kubo-Martin-Schwinger) state describes a thermal equilibrium for a quantum system at a specific inverse temperature. The paper establishes the existence of a unique KMS state for this spin system, which is central to the study of its thermodynamic properties.
- Regularity Conjecture
- The regularity conjecture suggests that certain states (like the KMS state) should be well-behaved, meaning their correlation functions decay in a specific way. This paper rigorously disproves this by showing a strict inequality involving relative entropy densities.
- Weak Gibbs State
- A weak Gibbs state is a specific type of quantum state that satisfies certain bounds related to the system's Hamiltonian and temperature. The nonregularity result demonstrates that the unique KMS state in this model does not satisfy these conditions, proving it is not weak Gibbs.
- Relative Entropy Density
- This quantity measures how much one probability distribution (the empty product state $ u$) differs from another (the KMS state $ ho$). The paper uses bounds on this density to show the nonregularity of the KMS state.
Terminology
Summary
A counterexample to the weak Gibbs and regularity conjectures is provided by a spin-one system on a square lattice with nearest-neighbor interaction and a uniform on-site field, demonstrating that its unique KMS state is nonregular and not weak Gibbs
The Nonregularity Result
The paper constructs a translation-invariant nearest-neighbor quantum spin system whose unique KMS state is nonregular The authors establish for a KMS state ω and an explicit translation-invariant state ν the strict inequality that disproves the regularity conjecture This is achieved by showing that lim sup k→∞ D(νΛLk ωΛLk) ΛLk ≤ pH(β) + 2βh - δ < pH(β) + 2βh
Model Definition and Properties
The model is defined on the square lattice Z squared, with a local Hilbert space C cubed and a Hamiltonian HZ involving hopping terms Fxy that turn one particle into two along the bond and two into one in the opposite direction The Hamiltonian conserves the dyadically weighted particle number CZ, which is invariant under HZ The existence of a unique KMS state ω for small inverse temperature β and arbitrary field strengths h is established
Proof Strategy via Gibbs Sampler
The proof proceeds by first proving uniqueness of the KMS state, then testing regularity against the empty product state ν The relative entropy density is defined as s(ν ω):= lim Z↗Zd D(νZ ωZ) Z = lim Z↗Zd trZ (log νZ − log ωZ) The authors use a Gibbs sampler to establish uniqueness and then bound the relative entropy between ω and ν using an imaginary-time return amplitude
Key Technical Steps in the Bound
The comparison between the empty-region probability and the return amplitude yields a bound on D(νΛL ωΛL) involving log ZΛL − log GΛ+L + O(L) By choosing a sufficiently large field h, the authors show that-log GΛ+Lk ≤ (2βh - δ)Λ+Lk This leads to the final contradiction where lim sup k→∞ D(νΛLk ωΛLk) ΛLk ≤ pH(β) + 2βh - δ < pH(β) + 2βh
Invariance of the KMS State
The authors prove that every translation-invariant state is KMS and that the unique KMS state ω is invariant under each limiting update Φx The invariance of the physical evolution implies that for every β-KMS state φ, φ(A) = φ(Tt(A)) → ω(A) as t → ∞ This demonstrates that omega is the unique KMS state
Conclusion on Weak Gibbsianity
Since the KMS state in the example is nonregular, it is also not weak Gibbs This result builds upon previous work that established failure of regularity in models with a unique KMS state The paper concludes that ω is the unique KMS state and every translate of ω is again KMS
References
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
-
Weak Gibbsianity detection: The improved system can determine whether a translation-invariant state is weak Gibbs by checking if there exist constants cZ >= 0 for finite lattice subsets Z such that e - cZ exp(-betaHZ) trZ exp(-betaHZ) <= omegaZ <= e cZ exp(-betaHZ) trZ exp(-betaHZ), and the limit of cZ / Z is zero.
-
Regularity conjecture falsification: The system can rigorously disprove the regularity conjecture for a unique KMS state by establishing the strict inequality lim inf Z↗Zd D(νZ∥ωZ) Z < -s(ν) + βeH(ν) + pH(β).
-
Model analysis capabilities: The AI can analyze spin-one systems on square lattices with nearest-neighbor interactions and uniform on-site fields to identify models that possess a unique KMS state but are not regular, which is crucial for understanding quantum breakdown contexts.
Abstract
We provide a counterexample to the weak Gibbs and regularity conjectures. The example is a spin-one system on the square lattice with a translation-invariant nearest-neighbor interaction. At high temperatures and high on-site field strengths, it has a unique KMS state that does not have the weak Gibbs or regularity properties.
Sources
- A Finite-Range, Translation-Invariant Counterexample to Weak Gibbsianity
- Rapid mixing of Gibbs samplers via quantum Dobrushin--Shlosman conditions
- Lieb-Robinson bounds for irreversible dynamics via adaptive Dyson expansion
Related papers
- Existence of a fluid phase for the two-dimensional one-component plasma
- The Huang-Yang conjecture for the low-density Fermi gas
- Anderson localisation in infinite and finite resonator crystals with long-range interactions
- Free energy and spectral edge of the SYK model
- Anderson localization for 1-d quasi-periodic Schr"odinger operators with degenerate weights
- Irregular SLE(4) martingales and isomonodromic deformations