Rigidity and equidistribution of random walks by diffeomorphisms near the conservative regime
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Mathematics. The day in mathematics: new theorems, new proofs, and what they open.
Ada: We're looking at the paper: "Rigidity and equidistribution of random walks by diffeomorphisms near the conservative regime".
Paper summary: Ada: So we’re looking at this paper, "Rigidity and equidistribution of random walks by diffeomorphisms near the conservative regime." The main point here is that if you have a random walk on a closed manifold M driven by a probability measure on diffeomorphisms with compact support, and it has these certain gap and pinching conditions, then there’s exactly one atom-free stationary probability measure. Emmy And this unique measure is the volume measure if the driving group preserves that volume. It also coincides with the volume measure in those cases where it does.
Kolya: I mean, so we have this statement about uniqueness for stationary measures and then equidistribution for almost every point, which is pretty strong. The paper claims that for every point x on M with an infinite group generated by its orbit under the random walk, the n-step distribution of that walk converges to this unique measure unless x gets stuck in a finite set.
Ada: Exactly. So we’re talking about proving this uniqueness and convergence under very specific conditions on the probability measure driving those diffeomorphisms. It really matters because it connects things like rigidity, which is often about fixed structures, with equidistribution, which is about how things spread out over the whole space. Emmy It sets up a framework where the volume measure is special and dominates the long-term behavior of these random walks when they are behaving well.
Kolya: What I find interesting is that they reduce their proof to three main phases: getting a dimensional estimate, then bootstrapping that to almost full dimension, and finally using spectral gaps for equidistribution. It’s a pretty layered argument. Ada Right, so it starts by showing the n-step distribution has positive dimension above some scale ρ of n which goes to zero as n goes to infinity. That sounds like a good starting point for building up complexity.
Emmy: And then they use that dimensional estimate to get almost full dimension using a local multislicing method, and finally they promote that high dimension to actual equidistribution by looking at the spectral gap of the convolution operator on a Sobolev space. It’s like they build it up layer by layer.
Conclusion: Ada: So wrapping up this discussion on "Rigidity and equidistribution of random walks by diffeomorphisms near the conservative regime," the authors are essentially showing that under those gap and pinching conditions, you get one unique atom-free stationary probability measure, which is the volume measure in volume-preserving settings. Emmy And for almost every point x on the manifold M that doesn't get trapped in a finite set, the n-step distribution of your random walk converges to this same unique measure.
Kolya: It’s about establishing that this convergence happens even when you’re dealing with these more general settings involving diffeomorphisms and random walks on closed manifolds. The implication is that for certain well-behaved systems, the volume measure isn't just *a* stationary measure; it's the only one that matters in the long run.
Ada: Right. It means we have a very specific target distribution to aim for when studying these random processes, and we know how to get there if the setup is right. Emmy So, this paper gives us a concrete toolset—the gap and pinching conditions—that allows us to predict not just what the stationary measure is, but also where the walk ends up after many steps.
Kolya: It’s a lot of technical machinery, but it’s powerful because it applies across several settings, like surfaces with bi-expanding random walks or even on cocompact lattice quotients of SO (two one) and SO (three one) <ref:2605.27008#pg1,on cocompact lattice quotients of SO (2 , 1>. The complexity is in making sure those specific conditions hold for the measure driving the walk.
Ada: That’s right. It shows that rigidity and equidistribution are linked in this context, meaning you can’t have one without the other when these conditions on the measure are met. And it really solidifies how fundamental these random dynamics are in geometry and analysis.
TIMOTHEE B ENARD, ZHIYUAN ZHANG
University of Paris 13 · Imperial College London
math.DS
Submitted: 2026-05-26
Updated: 2026-11-16
Comments: 50 pages. Corrected the stationary-measure regularity statement in the dissipative setting; clarified hypotheses and proofs, improved exposition, and updated references
License: http://creativecommons.org/licenses/by/4.0/
Importance score: 89/100
The gist: Rigidity and equidistribution of random walks by diffeomorphisms near the conservative regime proves that under certain conditions on a probability measure driving a random walk on a closed manifold…
Key concepts
- Rigidity
- This concept establishes that under specific conditions on a probability measure driving a random walk, the only possible atom-free stationary probability measure is the volume measure itself when the dynamics preserve volume. It means there is no other type of invariant distribution besides this unique one.
- Equidistribution
- This refers to how the distribution of a random walk's position spreads out over time. The paper shows that for points with infinite orbits under the measure, the n-step distribution converges to a specific unique measure, indicating that the walk becomes uniformly distributed according to this stationary measure.
- Gap and Pinching Conditions
- These are technical conditions placed on the probability measure driving the random walk. They quantify how much the dynamics expand or contract in different directions, ensuring that the random walk has enough 'room' to explore its space effectively without getting trapped in small regions.
Terminology
Summary
Rigidity and equidistribution of random walks by diffeomorphisms near the conservative regime proves that under certain conditions on a probability measure driving a random walk on a closed manifold, there exists exactly one atom-free stationary probability measure that coincides with the volume measure in the volume-preserving setting, and for almost every point, the n-step distribution converges to this unique measure
Main Results Summary
The paper establishes a rigidity result stating that for a random walk on a closed manifold driven by a probability measure on diffeomorphisms with compact support satisfying certain gap and pinching conditions, there is exactly one atom-free stationary probability measure
-
Rigidity: The volume measure vol is the only atom-free µ-stationary probability measure on M when Γµ preserves vol
-
Equidistribution: For every x in M with infinite Γµ-orbit, the n-step distribution µ∗n ∗ δx converges to ϒµ unless x is trapped in a finite µ-invariant set
Proof Strategy Reduction
The proof of Theorem 1.9 is reduced to three phases corresponding to Propositions 2. p. 8 9 to establishing three key results (namely Propositions 2. 11), corresponding to the
phases I, II, III mentioned in the introductory paragraph.Where the paper says it
(Phase I)
Phase I shows that for large enough n, the distribution µ∗n ∗ δx has positive dimension above a scale ρµ (n) > 0 where limn→+∞ ρµ (n) = 0
(Phase II)
Phase II bootstraps the dimensional estimate obtained in Phase I to almost full dimension using a local variant of the multislicing method from [BH, BH2]
(Phase III)
Phase III promotes high dimension to equidistribution by invoking the spectral gap of the convolution operator f→f∗µ on an appropriate Sobolev space
Key Properties and Conditions
The main theorem relies on a set of conditions on the measure µ0, including:
-
Gap and pinching conditions: µ0 has a (κ,J)-gap and is (η,J)-pinched
-
Expansion condition: µ−1 0 is (−κ/d)-expanding
-
Ergodicity of the Markov operator: Tµ−1 0 is totally ergodic in L2(M,vol)
The paper introduces concepts like (n0,κ,b)-gap and (n0,η,b0,b1)-pinching to quantify the expansion/contraction properties of the random walk
Dimensional Increment Mechanism
Proposition 2.10 shows that if certain conditions hold for µ and n is large enough, then µ∗n ∗ δx is (1−κ, Bρ, ρε)-robust
The proof of Proposition 6.2 uses the multislicing estimate provided by Theorem 5. p. 0 Initial dimension : Proof of Proposition 2. Benard-He’s Multislicing Theorem 27 ´
5. Dimensional increment : Proof of Proposition 2. Linearization at local scales 33
6.Where the paper says it
Where the paper says it
Equidistribution Mechanism
Proposition 7.2 proves that for a (1−ε,Bρ,τ)-robust measure ν, the Wasserstein distance between µ∗n ∗ ν and ϒµ is bounded by ρε + τ
This equidistribution relies on the spectral gap of the Markov operator Tµ on Hs(M) being smaller than 1−κ' and the non-concentration property of ν
Applications to Specific Settings
The results apply to several settings:
(Case I - Surface Perturbations)
For perturbations of surface maps, the expansion/coexpansion properties imply a 1-gap and coexpansion
(Case II - Torus Perturbations)
For linear random walks on the torus Td, the results follow from Bourgain-Furman-Lindenstrauss-Mozes [BFLM]
(Case III - Homogeneous Spaces)
For cocompact lattice quotients of SO(2,1) and SO(3,1), the paper shows that µ0 has a b-gap for every b in certain sets
Spectral Estimates
Theorem 7.3 establishes that the coexpanding property of µ0 guarantees an essential spectral gap for Tµ on Hs(M)
This is crucial for showing that the spectral projection at 1 is given by a probability measure ϒµ on M
Multislicing Theorem
Theorem 5.8 presents a multislicing statement adapted to the paper's needs, allowing one to show that for most parameters θ with respect to σ, the measure ν has improved dimension with respect to B Wθ ρtθ
Final Conclusion
The final step involves combining estimates from Lemma 4.9 and Proposition 4. p. 0 Expansion and coexpansion 20
4. Initial dimension : Proof of Proposition 2. Expansion and coexpansion 20
4. Initial dimension : Proof of Proposition 2.Where the paper says it
Where the paper says it
This leads to the conclusion that the measure µ∗n ∗ ν is (α +γ, Bρ1/2, ρε'/2)-robust for all x ∈ M with infinite Γµ-orbit
References
The paper cites numerous works including [BFLM], [BEFRH], and [DDZ].
[Bek] M. B. Bekka, On uniqueness of invariant means
[Bou] J. Bourgain, The discretized sum-product and projection theorems, J. p. 4 1 case (I), item (a) recovers the rigidity statement [BLOR2, Thm
B] and item (b) strenghtens the equidistribution statement [BLOR2, Thm C] by
removing the Cesaro average, hereby providing a positive answer to [BLOR2,
Question 1. We stress that [BLOR2] ultimately relies on drift arguments “a`
la Benoist-Quint/Eskin-Mirzakhani” (via [BRH]) while we employ a completely
different method rooted in Bourgain’s sum-product theorem.Where the paper says it
[BH] T. Benard and W. p. 0 Initial dimension : Proof of Proposition 2. Benard-He’s Multislicing Theorem 27 ´
5.Where the paper says it
[BHZ] T. Benard, W. He and H. Zhang, Khintchine dichotomy for self-similar measures, Journal of the American Mathematical Society (2026) 39, 587-623
[BLOR] A. Brown, H. Lee, D. Obata, Y. Ruan, Absolute continuity of stationary measures
[BLOR2] A. Brown, H. Lee, D. Obata, Y. Ruan, Absolute continuity of stationary measures for random surface dynamics
[BQ] Y. Benoist, J. p. 2 By means of Fourier
analysis and techniques from additive combinatorics, they are able to show rigidity (a) and equidistribution (b) as stated above. Soon after, Benoist-Quint [BQ2]
provide another proof of (a) for linear walks on the torus, making use of completely different tools inspired by the proof of Ratner’s rigidity theorems for unipotent flows. By means of Fourier
analysis and techniques from additive combinatorics, they are able to show rigidity (a) and equidistribution (b) as stated above. Soon after, Benoist-Quint [BQ2]
provide another proof of (a) for linear walks on the torus, making use of completely different tools inspired by the proof of Ratner’s rigidity theorems for unipotent flows. By means of Fourier
analysis and techniques from additive combinatorics, they are able to show rigidity (a) and equidistribution (b) as stated above. Soon after, Benoist-Quint [BQ2]
provide another proof of (a) for linear walks on the torus, making use of completely different tools inspired by the proof of Ratner’s rigidity theorems for unipotent flows. By means of Fourier
analysis and techniques from additive combinatorics, they are able to show rigidity (a) and equidistribution (b) as stated above. Soon after, Benoist-Quint [BQ2]
provide another proof of (a) for linear walks on the torus, making use of completely different tools inspired by the proof of Ratner’s rigidity theorems for unipotent flows.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
[BQ2] Y. Benoist, J.-F. Quint, Stationary measures and closed invariants on homogeneous spaces
[BQ3] Y. Benoist, J.-F. Quint, Stationary measures and invariant subsets of homogeneous spaces II, Journal of the American Mathematical Society 26 (2013), 659-734
[BQ4] Y. Benoist, J.-F. Quint, Stationary measures and invariant subsets of homogeneous spaces III, Annals of Mathematics 178 (2013), 1017-1059
[BRH] A. Brown, F. Rodriguez Hertz, Measure rigidity for random dynamics on surfaces and related skew products, J. p. 2 Regarding non-homogeneous settings, Benoist-Quint’s approach has been influential in the groundbreaking work of Eskin-Mirzakhani [EM], on rigidity properties of the SL2(R)-action on moduli space of translation surfaces, and in the
landmark work of Brown-Rodriguez Hertz [BRH] dealing with random walks by
diffeomorphisms on closed surfaces. The latter shows that any ergodic stationary
measure for a random walk on a surface falls in one of three categories: either it is
uniform on a finite invariant set, or the stable direction of the walk is non-random,
or the measure is SRB meaning it is leafwise absolutely continuous for a notion of
unstable foliation.Where the paper says it
[CD] S. Cantat and R. Dujardin, Dynamics of automorphism groups of projective surfaces: classification, examples and outlook, arXiv: 2310. p. 2 Benoist-Quint’s method has the advantage to extend to arbitrary lattice
quotients of Lie groups provided the measure driving the walk satisfies a certain
semisimplicity assumption [BQ2, BQ3]. In [BQ4], classification of stationary measures allow Benoist-Quint to derive a weak version of (b), in the sense that convergence only holds in Cesaro average:
1
n
n−1
∑
k=0
µ
∗k
∗ δx → vol.Where the paper says it
[CG] J-P. Conze, Y. Guivarc’h, Ergodicity of group actions and spectral gap, applications to random walks and Markov shifts, Discrete and Continuous Dynamical Systems, 2013, 33(9):4239-4269
[DD] J. DeWitt and D. p. 5 In the above perturbed context, Dolgopyat-Krikorian [DK] show that for
d ∈ 2N even, the volume measure is the only absolutely continuous µ-stationary
probability on S
d
. This result is extended to all d ≥ 2 in the recent work of DeWitt,
Dolgopyat and the second author [DDZ].Where the paper says it
[DDZ] J. DeWitt, D. Dolgopyat and Z. p. 5 5 follows from Ito-Kawada’s theorem ˆ
[IK]. In the above perturbed context, Dolgopyat-Krikorian [DK] show that for
d ∈ 2N even, the volume measure is the only absolutely continuous µ-stationary
probability on S
d
. Proofs of Theorem 1. 7 and applications 40 1
arXiv:2605.Where the paper says it
Where the paper says it
[DK] D. Dolgopyat and R. p. 5 5 follows from Ito-Kawada’s theorem ˆ
[IK]. In the above perturbed context, Dolgopyat-Krikorian [DK] show that for
d ∈ 2N even, the volume measure is the only absolutely continuous µ-stationary
probability on S
d
. We will comment
on this in Remark 8. We may also apply our method to study random walks on the sphere M = S
d
driven by random diffeomorphisms close to rotations.Where the paper says it
Where the paper says it
[ES] R. Elliot Smith, TBA, (2026)
[EM] A. Eskin and M. p. 2 In [BQ4], classification of stationary measures allow Benoist-Quint to derive a weak version of (b), in the sense that convergence only holds in Cesaro average:
1
n
n−1
∑
k=0
µ
∗k
∗ δx → vol. Regarding non-homogeneous settings, Benoist-Quint’s approach has been influential in the groundbreaking work of Eskin-Mirzakhani [EM], on rigidity properties of the SL2(R)-action on moduli space of translation surfaces, and in the
landmark work of Brown-Rodriguez Hertz [BRH] dealing with random walks by
diffeomorphisms on closed surfaces. Regarding non-homogeneous settings, Benoist-Quint’s approach has been influential in the groundbreaking work of Eskin-Mirzakhani [EM], on rigidity properties of the SL2(R)-action on moduli space of translation surfaces, and in the
landmark work of Brown-Rodriguez Hertz [BRH] dealing with random walks by
diffeomorphisms on closed surfaces. The latter shows that any ergodic stationary
measure for a random walk on a surface falls in one of three categories: either it is
uniform on a finite invariant set, or the stable direction of the walk is non-random,
or the measure is SRB meaning it is leafwise absolutely continuous for a notion of
unstable foliation.Where the paper says it
Where the paper says it
[EPZ] A. Eskin, R. Potrie and Z. Zhang, Geometric properties of partially hyperbolic measures and applications to measure rigidity, arXiv: 2302. p. 0 Proofs of Theorem 1. 7 and applications 40 1
arXiv:2605.Where the paper says it
[FM] D. Fisher and G. Margulis, Almost isometric actions, property (T), and local rigidity, Invent. p. 3 Noting that rigidity (a) follows from
equidistribution (b), the proof is divided in three steps: (Phase I) for large enough
n, we show the distribution µ
∗n ∗ δx has positive dimension above scale ρµ (n) > 0
where limn→+∞ ρµ (n) = 0. (Phase II) We bootstrap positive dimension to almost
full dimension using a local variant of the multislicing method from [BH, BH2].Where the paper says it
[He] W. He, Orthogonal projections of discretized sets, J. p. 6 Informally, a measure µ on G has a
b-gap if at every point x, the b-th and (b+1)-th smallest Lyapunov exponents are
separated by a constant κ = κ(µ) > 0. We formulate this condition by asking that,
for the action of g ↷ M where g is chosen by µ (or an iterate), the strongest “contraction” along any (d −b)-dimensional tangent subspace dominates the strongest
“expansion” along its b-dimensional orthogonal complement.Where the paper says it
[He2] W. He, Random walks on linear groups satisfying a Schubert condition, Isr. p. 2 Soon after, Benoist-Quint [BQ2]
provide another proof of (a) for linear walks on the torus, making use of completely different tools inspired by the proof of Ratner’s rigidity theorems for unipotent flows. Benoist-Quint’s method has the advantage to extend to arbitrary lattice
quotients of Lie groups provided the measure driving the walk satisfies a certain
semisimplicity assumption [BQ2, BQ3].Where the paper says it
[Hen] H. Hennion, Sur un theor´eme spectral et son application aux noyaux lipschitziens, Proc. Amer. Math. Soc. 118(2), (1993) 627-634
[Hir] M. Hirsch, Differential Topology, Graduate texts in mathematics, 33 (1997), Berlin, New York: Springer-Verlag
[IK] Y. Kawada and K. p. 5 In the linear regime µ = µ0, Theorem 1. 5 follows from Ito-Kawada’s theorem ˆ
[IK]. 5 follows from Ito-Kawada’s theorem ˆ
[IK]. In the above perturbed context, Dolgopyat-Krikorian [DK] show that for
d ∈ 2N even, the volume measure is the only absolutely continuous µ-stationary
probability on S
d
.Where the paper says it
Where the paper says it
[KK] C. Kogler and W. Kim, Effective density of non-degenerate random walks on homogeneous spaces, Int. p. 7 7 allows for any values of the Lyapunov
exponents, however they need to have multiplicity one. This rules out applicability
to perturbation of linear random walks on complex homogeneous spaces, such as
quotients of SO(3,1) (which is isogeneous to SL2(C)).Where the paper says it
[KL] G. Keller and C. Liverani., Stability of the spectrum for transfer operators, Ann. p. 5 We may also apply our method to study random walks on the sphere M = S
d
driven by random diffeomorphisms close to rotations. Here we rely on a rigidity
theorem of [DK] claiming that a perturbation of a finitely supported measure on
SO(d +1) spanning a dense subgroup is either conjugated to a measure on SO(d +
1) or has simple Lyapunov spectrum. • Vector spaces. We introduce some notations related to a finite dimensional real
vector space V equipped with a norm induced by an inner product.Where the paper says it
Where the paper says it
[LMW] E. Lindenstrauss, A. p. 5 In the linear regime µ = µ0, Theorem 1. 1 case (II) is a consequence of BourgainFurman-Lindenstrauss-Mozes [BFLM], see aslo [BQ2].Where the paper says it
[M] J. Moser, On the volume elements on a manifold, Trans. p. 1 In the dissipative setting, we still require µ to be weak-∗ close to a volume-preserving
probability measure (conservative case). Note this allows supp(µ)to contain highly
volume-disrupting elements, although with small mass.Where the paper says it
[Pot] R. Potrie, A remark on uniform expansion, Rev. p. 2 The latter shows that any ergodic stationary
measure for a random walk on a surface falls in one of three categories: either it is
uniform on a finite invariant set, or the stable direction of the walk is non-random,
or the measure is SRB meaning it is leafwise absolutely continuous for a notion of
unstable foliation. Building on techniques from Tsujii [Tsu], Brown-Lee-ObataRuan [BLOR, BLOR2] are able to promote this SRB conclusion to absolute continuity with respect to volume under certain expansion conditions.Where the paper says it
[Rod] M. Roda, Classifying Hyperbolic Ergodic Stationary Measures on Compact Complex Surfaces with Large Automorphism Groups, arXiv: 2410.18350
Pruning redundant reasoning tokens
The paper introduces the strategy of reducing Theorem 1.9 to three key results (Propositions 2. p. 8 In Section 2, we reduce Theorem 1. 9 to establishing three key results (namely Propositions 2.Where the paper says it
How the training signal is built
The dimensional increment is achieved through an iterative application of Proposition 2.8
The proof of Proposition 4.3 borrows from [BH, Section 5] which deals with the context of Zariski-dense random walks on semisimple homogeneous spaces
Where the gains come from
The dimensional increment is guaranteed for the interval of scales [ρC/2k,ρ1/(2kC)] and a trash component of the form Oε,C,k(ρε/4C)
The equidistribution in Phase III relies on the spectral gap of the convolution operator by µ on a suitable Sobolev space
The final estimate shows that µ∗n ∗ ν(f) ≲µ,κ′ ρ2ε +ρ1/2 and taking ρ≪µ,κ′,ε 1, the result follows
This is achieved by combining estimates from Lemma 4.9 and Proposition 4.3
The final proof of Theorem 1.9 shows that ϒµ is the unique atom-free µ-stationary probability measure
This is achieved by combining Proposition 2.10 and Proposition 2. p. 0 Expansion and coexpansion 20
4. Initial dimension : Proof of Proposition 2.Where the paper says it
The proof of Theorem 1.9 relies on Lemma 3. p. 4 1 case (I), item (a) recovers the rigidity statement [BLOR2, Thm
B] and item (b) strenghtens the equidistribution statement [BLOR2, Thm C] by
removing the Cesaro average, hereby providing a positive answer to [BLOR2,
Question 1. We stress that [BLOR2] ultimately relies on drift arguments “a`
la Benoist-Quint/Eskin-Mirzakhani” (via [BRH]) while we employ a completely
different method rooted in Bourgain’s sum-product theorem.Where the paper says it
This is achieved by combining Theorem 7.4 and Theorem 1. p. 0 Initial dimension : Proof of Proposition 2. Benard-He’s Multislicing Theorem 27 ´
5. Initial dimension : Proof of Proposition 2. Benard-He’s Multislicing Theorem 27 ´
5.Where the paper says it
Where the paper says it
The proof of Theorem 1.1, case I relies on Lemma 3. p. 4 1 case (I), item (a) recovers the rigidity statement [BLOR2, Thm
B] and item (b) strenghtens the equidistribution statement [BLOR2, Thm C] by
removing the Cesaro average, hereby providing a positive answer to [BLOR2,
Question 1. We stress that [BLOR2] ultimately relies on drift arguments “a`
la Benoist-Quint/Eskin-Mirzakhani” (via [BRH]) while we employ a completely
different method rooted in Bourgain’s sum-product theorem.Where the paper says it
This is achieved by combining Theorem 7.4 and Theorem 1.7
The proof of Theorem 1.1, case II relies on Lemma 8. p. 4 1 case (I), item (a) recovers the rigidity statement [BLOR2, Thm
B] and item (b) strenghtens the equidistribution statement [BLOR2, Thm C] by
removing the Cesaro average, hereby providing a positive answer to [BLOR2,
Question 1. We stress that [BLOR2] ultimately relies on drift arguments “a`
la Benoist-Quint/Eskin-Mirzakhani” (via [BRH]) while we employ a completely
different method rooted in Bourgain’s sum-product theorem.Where the paper says it
The proof of Theorem 1.1, case III relies on Lemma 8. p. 4 1 case (I), item (a) recovers the rigidity statement [BLOR2, Thm
B] and item (b) strenghtens the equidistribution statement [BLOR2, Thm C] by
removing the Cesaro average, hereby providing a positive answer to [BLOR2,
Question 1. We stress that [BLOR2] ultimately relies on drift arguments “a`
la Benoist-Quint/Eskin-Mirzakhani” (via [BRH]) while we employ a completely
different method rooted in Bourgain’s sum-product theorem.Where the paper says it
The proof of Proposition 3.6 relies on Lemma 3. p. 4 1 case (I), item (a) recovers the rigidity statement [BLOR2, Thm
B] and item (b) strenghtens the equidistribution statement [BLOR2, Thm C] by
removing the Cesaro average, hereby providing a positive answer to [BLOR2,
Question 1. We stress that [BLOR2] ultimately relies on drift arguments “a`
la Benoist-Quint/Eskin-Mirzakhani” (via [BRH]) while we employ a completely
different method rooted in Bourgain’s sum-product theorem.Where the paper says it
This is achieved by combining Theorem 1.9 and Lemma 3. p. 4 This follows from [Pot] and the proof therein,
(2) It is equivalent to the 1-gap property and to the coexpanding property (see
Definitions 1. 11 and Lemma 3. This follows from [Pot] and the proof therein,
(2) It is equivalent to the 1-gap property and to the coexpanding property (see
Definitions 1. 11 and Lemma 3.Where the paper says it
Where the paper says it
This is achieved by combining Theorem 1.9 and Lemma 3.
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
-
Improved ability to prove rigidity for random walks on general closed manifolds by classifying stationary measures and establishing equidistribution properties under gap and pinching conditions.
-
Enhanced capacity to handle non-homogeneous settings by developing methods that allow for the existence of a unique atom-free stationary probability measure even when the driving measure is not volume-preserving, leading to a full dimensional result in the dissipative setting.
-
Increased precision in characterizing singular values of random walks by formulating gap and pinching conditions, allowing for quantitative estimates on the separation between Lyapunov exponents as a function of time steps n.
-
Capability to perform dimensional increment proofs by iteratively applying multislicing methods, showing how the distribution mu star n star delta x ultimately reaches high dimension above any prescribed scale ρ0.
-
Stronger ability to establish equidistribution results by employing spectral gap estimates for the Markov operator on Sobolev spaces, leading to quantitative bounds on the Wasserstein distance between the n-step distribution and a stationary measure.
-
Advanced understanding of geometric properties of random walks by applying projection theorems (subcritical and supercritical) in a multislicing framework to show that most random boxes satisfy improved dimensional properties with respect to the underlying measure ν.
Abstract
We consider a random walk on a closed manifold M driven by a probability measure μ on the set of C squared-diffeomorphisms. Provided μ has compact support, satisfies certain gap, pinching, and ergodicity conditions, and is weak- * close to volume-preserving, we prove M carries exactly one atom-free stationary probability measure Υ μ. It has finite high-dimensional Riesz energy and coincides with volume in the volume-preserving setting. Moreover, for every x in M, the n-step distribution μ*n*δ x converges to Υ μ unless x is trapped in a finite μ-invariant set. Our result applies to a variety of situations: bi-expanding random walks on surfaces, non-linear perturbations of Zariski-dense random walks on the torus T d, on cocompact lattice quotients of SO(2,1), SO(3,1), or on the sphere S d.
Sources
- Uniform spectral gaps, non-abelian Littlewood-Offord and anti-concentration for random walks
- Measure rigidity for generalized u-Gibbs states and stationary measures via the factorization method
- Effective equidistribution of random walks on simple homogeneous spaces
- Absolute continuity of stationary measures
- Absolute continuity of stationary measures for random surface dynamics
- Dynamics of automorphism groups of projective surfaces: classification, examples and outlook
- Conservative Coexpanding on Average Diffeomorphisms
- Expanding on average diffeomorphisms of surfaces: exponential mixing
- Ergodicity of (co)expanding on average random dynamical systems
- Geometric properties of partially hyperbolic measures and applications to measure rigidity
- Classifying Hyperbolic Ergodic Stationary Measures on Compact Complex Surfaces with Large Automorphism Groups
Related papers
- Proof of the singularity conjecture for discrete subgroups of SL 2(R)
- Multiple ergodic averages for commuting multiplicative actions
- A finite Livshits theorem and local length spectrum rigidity
- Topological Detection of Hopf Bifurcations via Persistent Homology: A Functional Criterion from Time Series
- A hyperbolic area proof of the Kahn-Lyubich Covering Lemma and its generalization
- A Dynamical Fekete-Szeg Theorem