REVIEW 2 major objections 3 minor 11 references
Non-conformality of large deviations of moving average of the random walk in strongly mixing environment
T0 review · 2 major / 3 minor · reviewed 2026-08-07 · deepseek-v4-flash
Pith's one-line read In any genuinely random strongly mixing environment, quenched and annealed large deviations for a random walk always disagree at some interior velocity.
desk verdict Plausible extension of Yilmaz to mixing environments, but the proof's key inequality (4.2) has the wrong sign, so the non-conformality theorem is not established. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The central object is the auxiliary Q_z-walk, whose transition probabilities u_z(e) are chosen so that the walk has drift z and the environment ratios ξ factor through its increments through the identity (3.1). The proof also uses a stopping-time sequence $τ^{{(L)}}$_k that creates long blocks of steps in a fixed direction ℓ, and a separation lemma from the author's earlier preprint that controls the dependence between blocks in strongly mixing fields. The strict inequality between rate functions emerges from Jensen's inequality on the logarithm of the quenched expectation of a block product of environment ratios, which is strict exactly because the environment is non-deterministic.
What would settle it
A concrete check: in d=1, take an i.i.d. environment with two possible values of ω(0,1), compute I_a and I_q numerically, and see if they are equal at some x∈(-1,1); if they are, Theorem 2.1 is false, and if the separation lemma fails on any (SMX)_{C,g} example, the proof breaks.
Extended reading notes
Core claim
Theorem 2.1 states that for any d≥1 and any uniform ellipticity constant κ>0, every environment law P satisfying the strongly mixing condition (SMX)_{C,g} with dis(P)>0 produces an interior point x of the velocity surface D={|x|_1≤1} where I_a(x)<I_q(x). The proof constructs an auxiliary random walk Q_z with drift z, couples its increments to the environment ratios ξ(x,e)=ω(x,e)/E[ω(x,e)], and uses a separation lemma for correlated fields to compare quenched and annealed free energies along a stopping-time sequence. Jensen's inequality applied to a block of the coupled walk yields a strict gap at a vertex direction ℓ; continuity of the rate functions then transfers the strict inequality to an interior point. The result is stated in the strongest form: non-conformality holds for every non-singleton strongly mixing environment, with no smallness condition on the disorder.
Load-bearing premise
The proof depends on the separation lemma and the pointwise large-deviation estimates from the author's earlier preprint being valid for the whole (SMX)_{C,g} class; if any of those lemmas fails or needs extra conditions, the conclusion does not follow.
Editorial extensions
If this is right
- The known low-disorder conformality of I_a and I_q on compact sets cannot be extended to the whole velocity surface: a gap always remains at some interior point.
- For every non-singleton strongly mixing environment in any dimension d≥1, there are velocities at which the typical fluctuations of the walk are described differently by the quenched and annealed averages.
- The result generalizes Yilmaz's i.i.d. non-conformality to dependent random fields under strong mixing, removing the i.i.d. restriction and covering all dimensions.
- Disorder level does not control global non-conformality: even as dis(P)→0 (but positive), the two rate functions fail to agree somewhere in the interior of the domain.
Reading between the lines
- The proof's mechanism implies the edge-vertex set ∂D^{d-2} is the natural birthplace of non-conformality: the strict gap is produced at a vertex and pushed inward by continuity; a quantitative lower bound on the gap in terms of dis(P) and the mixing constants would be a natural next step.
- Because the paper leans on unproduced lemmas from the author's earlier preprint, the theorem's scope is only as wide as those lemmas; if they require extra hypotheses, the result may shrink to a subclass of (SMX)_{C,g} environments.
- The same non-conformality question for the weaker (SM)_{C,g} or (SMG)_{C,g} mixing conditions is left open; one could try to adapt the stopping-time construction to those settings.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper claims that for a random walk in a strongly mixing random environment satisfying (SMX)_{C,g}, the quenched and annealed large-deviation rate functions always differ at some interior point of the velocity surface whenever the environment is genuinely random (dis(P)>0). This is presented as an extension of Yilmaz's result from i.i.d. environments to dependent strongly mixing environments. The proof introduces an auxiliary Q^z-walk with tilted increments, identifies quenched and annealed point-to-point free energies through the auxiliary transition, and then compares the constrained quenched free energy with the annealed point-to-point free energy using a separation lemma and Jensen's inequality. The main conclusion is that the strict inequality I_q(x)<I_a(x) holds at some interior point.
Significance. If the result were sound, it would be a notable extension of a classical non-conformality result to dependent environments, and it would clarify the limits of the low-disorder conformality theorems previously established by the author. The construction of the auxiliary Q^z-walk and the change-of-measure identities are elegant, and the Jensen step is conceptually clear. However, the proof as written is not convincing: the central estimate (4.2) has the wrong inequality direction, and the argument relies on several lemmas from the author's unpublished preprint [3] without statements or proofs. The paper is too short to verify the many structural claims imported from [3], and the key quantitative step is not merely unproven but appears to be false in simple homogeneous examples. Thus the claimed theorem is not established by this manuscript.
major comments (2)
- [Section 4, Eq. (4.2)] The inequality in (4.2) has the wrong direction. The passage from the constrained event {⟨Z_N,ℓ⟩>(1−δ)N} to (1−δ) times the all-ℓ point-to-point event asserts that the constrained free energy is bounded above by (1−δ) times the point-to-point free energy plus O(δ). This is not justified and is generally false. Since the rate function I is convex, for any velocity v=(1−δ)ℓ+δw in the event one has I(v) ≤ (1−δ)I(ℓ)+δI(w) ≤ (1−δ)I(ℓ)+O(δ), so the logarithm of the constrained probability is at least −(1−δ)I(ℓ)+O(δ), the reverse of the asserted bound. The sentence 'the number of paths … is e^{O(ε)N}' neglects the exponential weights of the paths and the slope of the rate function. In the homogeneous one-dimensional Bernoulli example with p_+=0.6 and δ=0.01, the constrained free energy is −I(0.99)≈−0.4589, while (1−δ)I(1)≈0.5057, so the claimed ≤ direction fails. Consequently (4.2) cannot be used to derive the strict inequality I_q(ℓ)>I_a(ℓ) in (4.4).
- [Sections 2 and 4] The proof of Theorem 2.1 relies on [3, Theorem 2.1], [3, Lemma 4.1], and [3, Lemma 4.4] without stating their hypotheses or supplying proofs. These are inputs specific to the same model and are taken from an unpublished preprint by the author. The paper therefore does not provide a self-contained verification of the main theorem. If any of these statements requires additional conditions or is flawed, the conclusion of Theorem 2.1 does not follow. The author should either prove these results in the present paper or quote them with full statements and a clear indication of all assumptions used.
minor comments (3)
- [Section 2] The text contains typographical errors, e.g., 'hava revealed' should be 'have revealed', and several formulas lack spacing around operators.
- [Proof of Theorem 2.1] The notation 'supportP' is used without definition; it presumably means the support of the environment law P and should be defined explicitly.
- [Section 3] The letter κ is used both for the ellipticity constant and for the coin-flip probability к=к(κ); this overloaded notation is confusing and should be changed.
Circularity Check
Theorem 2.1's proof leans on the author's prior preprint [3] for the LDP and separation lemmas; the self-citation is load-bearing, though the Jensen step supplies independent content if those lemmas hold.
-
self citation load bearing
[Section 2 (LDP quoted from [3]) and Section 4, proof of Theorem 2.1, Eqs. (4.1)-(4.3)]
"In [3, Theorem 2.1], when P satisfies (SMX)_{C,g} the RWRE verifies the following quenched and annealed LDP on Z^d for any d≥1: ... where the last line follows from Lemma [3, Lemma 4.4]. ... Invoking a separation lemma for the stopping sequence (τ_k^{(L)})_{k≥1} via [3, Lemma 4.1], we have ..."
The proof of Theorem 2.1 takes as inputs, without reproduction or independent verification, (i) the existence, convexity, and continuity of I_a and I_q from [3, Theorem 2.1], and (ii) the separation lemmas [3, Lemmas 4.1 and 4.4] that convert full-path probabilities into block probabilities in (4.1)-(4.3). These imported results are from the author's own earlier preprint. The final conclusion I_a(x)<I_q(x) is obtained by applying Jensen's inequality to quantities whose very definition and block factorization come from [3]; if [3] were false, Theorem 2.1 would not follow. This makes the self-citation load-bearing rather than merely bibliographic.
full rationale
The only circularity-type issue in the paper is the repeated, load-bearing reliance on the author's own preprint [3] for the quenched/annealed LDP and for two separation lemmas. The main theorem is not derived from first principles in the present paper: Eqs. (4.1) and (4.3) explicitly send the reader to [3, Lemma 4.4] and [3, Lemma 4.1], and the existence of the rate functions is quoted from [3, Theorem 2.1]. Since [3] is an arXiv preprint by the same author that is not machine-checked, code-reproduced, or independently verified, it does not qualify as independent evidence under the review rules. The central inequality I_q(ℓ)>I_a(ℓ), however, is not equivalent to any input: the change of measure, stopping-time construction, and strict Jensen inequality (4.4) are carried out in this paper and would yield genuine content conditional on [3]. The reader's alleged sign error in (4.2) is a correctness objection, not a circularity objection, so it does not increase the circularity score. Score 4 reflects the substantial self-citation burden without finding that the derivation is definitionally circular.
Assumptions & free parameters
free parameters (3)
- C_z^P =
solution of f(C)=1 in Section 3
- к(κ) =
d^{-1} ∧ inf_{e∈V} u_z(e)
- u_z(e) =
u_z(e)=(⟨z,e⟩+sqrt(...))/2
assumptions (6)
- domain assumption Uniform ellipticity: P(ω(x,e)≥κ)=1 for all x,e
- domain assumption Strong mixing condition (SMX)_{C,g} and r-Markovian property
- domain assumption Existence and form of quenched and annealed LDP from [3, Theorem 2.1]
- domain assumption Separation lemma [3, Lemma 4.1] and Lemma [3, Lemma 4.4]
- standard math Path-counting estimate: number of paths in event is e^{O(ε)N}
- standard math Jensen's inequality
invented entities (3)
-
Q^z-walk (Z_n)
-
Auxiliary transition ψ_k
-
Stopping sequence τ_k^{(L)}
Cite this review
Pith. "Pith review of Non-conformality of large deviations of moving average of the random walk in strongly mixing environment." pith.science (2026). https://pith.science/paper/DOB6TV3V
@misc{pith2026250601316,
author = {Pith},
title = {Pith review of: Non-conformality of large deviations of moving average of the random walk in strongly mixing environment},
year = {2026},
howpublished = {\url{https://pith.science/paper/DOB6TV3V}},
note = {Machine review of arXiv:2506.01316}
}
read the original abstract
The quenched and annealed large deviations of the random walk in random environment are shown to conform on any compact set whenever the level of disorder is sufficiently low. In this work, we show that these two large deviations always disagree at some interior point of the natural domain of the random walk in strongly mixing environment, regardless of the level of disorder.
Reference graph
Works this paper leans on
-
[3]
J. Chen. Velocity surface disorder of large deviation rate functions of the random walk in strongly mixing environment. arXiv preprint,arXiv: 2409.06581, 2024
work page Pith review arXiv 2024
-
[1]
S. Alili. Asymptotic behavior for random walks in random environments.J. Appl. Probab.,36: 334–349, 1999
work page 1999
- [2]
- [4]
-
[5]
A. Drewitz, A. Ram´ ırez. Selected Topics in Random Walks in Random Environment.Topic. Percolat. Disorder. Syst., Springer Proceeding. Math. Stat.,69, 2014
work page 2014
- [6]
-
[7]
S. M. Kozlov. The averaging method and walks in inhomogeneous environments.Russian Math. Survey.,40 (2): 73–145, 1985
work page 1985
-
[8]
F. Rassoul-Agha, T. Sepp¨ al¨ ainen. Quenched point-to-point free energy for random walks in random potentials.Probab. Theor. Relat. Fields,158: 711–750, 2014
work page 2014
Show all 11 references
-
[9]
Y. Sinai. Lorentz gas and random walks.Math. Problem. Theor. Phys., Lect. Note. Phys.,153, 12–14, 1982
1982
-
[10]
D. E. Temkin. The theory of diffusionless crystal growth.J. Cryst. Growth,5 (3): 193–202, 1969
1969
-
[11]
A. Yilmaz. Equality of averaged and quenched large deviations for random walks in random environments in dimensions four and higher.Probab. Theor. Relat. Fields,149: 463–491, 2011. 5
2011
Reviewed August 7, 2026 · model on record in the stance chip above.
Discussion (0). Sign in to comment.