pith. sign in

arxiv: 2504.16532 · v3 · submitted 2025-04-23 · 🧮 math.DS

Optimal linear response for Anosov diffeomorphisms

Gary Froyland , Maxence Phalempin This is my paper

Pith reviewed 2026-05-22 18:41 UTC · model grok-4.3

classification 🧮 math.DS
keywords Anosov diffeomorphismsSRB measureslinear responseoptimal perturbationFourier coefficientstransfer operatornumerical approximation
0
0 comments X

The pith

The optimal perturbation maximizing the linear response of an SRB measure to a fixed observation is unique and given explicitly by its Fourier coefficients.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper develops a method to choose the best infinitesimal change to an Anosov diffeomorphism so that the long-term average of a chosen observation function shifts by the largest possible amount under the SRB measure. It proves this best perturbation is unique when the response satisfies a non-degeneracy condition and supplies explicit formulas for every Fourier coefficient of the corresponding vector field. A numerical procedure that truncates these Fourier series is shown to converge to the true optimum. The construction therefore supplies a concrete way to steer the statistics of a hyperbolic system with arbitrarily small adjustments to the map itself.

Core claim

For a fixed observation function c, the optimising perturbation dot T of an Anosov diffeomorphism T is unique for non-degenerate response functions. Explicit expressions are derived for the Fourier coefficients of dot T. An efficient Fourier-based numerical scheme approximates the optimal vector field dot T with a proof of convergence. Numerical examples demonstrate the localisation of SRB measures using these small, optimally selected perturbations.

What carries the argument

The linear response of the SRB measure expressed through the transfer operator, optimized over admissible perturbations to produce a vector field whose Fourier series is given in closed form.

If this is right

  • The optimising perturbation is unique whenever the response function is non-degenerate.
  • Every Fourier coefficient of the optimal vector field admits an explicit expression.
  • Truncation of the Fourier series yields a convergent numerical approximation to the optimal perturbation.
  • Small perturbations chosen this way can concentrate the SRB measure inside prescribed regions of phase space.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same Fourier representation could be used to optimize responses over finite time intervals rather than the infinite-time SRB average.
  • The method provides a practical route to numerical control of invariant measures in systems that admit linear response.
  • Extensions to families of observation functions would follow by treating the response as a vector-valued optimization problem.

Load-bearing premise

The SRB measure responds linearly to small changes in the map, so that the first-order change in its expectation of any observation is captured by a derivative of the transfer operator.

What would settle it

A concrete numerical test that finds another perturbation of the same size producing a strictly larger shift in the SRB expectation than the computed optimum, or that produces two distinct perturbations achieving exactly the same maximal shift in a non-degenerate case, would falsify the uniqueness and optimality claims.

Figures

Figures reproduced from arXiv: 2504.16532 by Gary Froyland, Maxence Phalempin.

Figure 1
Figure 1. Figure 1: Optimal vector field computed with n = 32 Fourier modes along each coordinate direction, using a Sobolev norm scaling factor γ = 0.02. The red lines are a segment of the stable manifold of the fixed point at the origin. A uniform scaling of the length of the vectors in this figure has been made for improved visualisation [PITH_FULL_IMAGE:figures/full_fig_p029_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: SRB measure estimate of Tδ = T0 + δ · T˙ for some positive δ, where the average norm of δ · T˙ is approximately 2% of the domain diameter. Note the concentration of mass about the fixed point near the origin. for ∆ = 0.01; see [4] for a proof that T0 is Anosov. The map T0 has two hyperbolic period-two orbits, and we focus on the period-two orbit (p1, p2) :≈ {(0.1796, 0.4023),(0.7877, 0.5852)}. We select th… view at source ↗
Figure 3
Figure 3. Figure 3: Optimal vector field computed with n = 32 Fourier modes along each coordinate direction, using a Sobolev norm scaling factor γ = 0.02. The blue background is an estimate of the SRB measure of the nonlinear Anosov map T0. A uniform scaling of the length of the vectors in this figure has been made for improved visualisation. points. In contrast to [PITH_FULL_IMAGE:figures/full_fig_p031_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Estimate of the SRB measure of Tδ = T0 + δ · T˙ for some positive δ, where the average norm of δ · T˙ is approximately 1.2% of the domain diameter. Note the strong concentration of mass about the period-two orbit. 7 Acknowledgements The research of GF and MP is supported by an Australian Research Council Laureate Fellowship (FL230100088). The authors acknowledge the helpful comments of an anonymous referee… view at source ↗
read the original abstract

It is well known that an Anosov diffeomorphism $T$ enjoys linear response of its SRB measure with respect to infinitesimal perturbations $\dot{T}$. For a fixed observation function $c$, we develop a theory to optimise the response of the SRB-expectation of $c$. Our approach is based on the response of the transfer operator on the anisotropic Banach spaces of Gou\"ezel--Liverani. We prove that the optimising perturbation $\dot{T}$ is unique for non-degenerate response functions and provide explicit expressions for the Fourier coefficients of $\dot{T}$. We develop an efficient Fourier-based numerical scheme to approximate the optimal vector field $\dot{T}$, along with a proof of convergence. The utility of our approach is illustrated in two numerical examples, by localising SRB measures with small, optimally selected, perturbations.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 3 minor

Summary. The paper develops a framework for optimizing the linear response of SRB measures for Anosov diffeomorphisms T under infinitesimal perturbations dot T, for a fixed observation function c. Working in the Gouëzel-Liverani anisotropic Banach spaces, it establishes uniqueness of the optimizing dot T for non-degenerate response functions, derives explicit Fourier coefficients for dot T, constructs a Fourier-based numerical approximation scheme with a convergence proof, and illustrates the method on two examples that localize SRB measures via small optimal perturbations.

Significance. If the central claims hold, the work supplies a concrete optimization procedure for controlling SRB expectations via linear response, together with explicit formulas and a provably convergent numerical method. The reliance on standard transfer-operator techniques on Gouëzel-Liverani spaces, combined with the parameter-free character of the uniqueness result for non-degenerate cases and the reproducible Fourier scheme, constitutes a clear advance in the quantitative study of linear response for hyperbolic systems.

major comments (2)
  1. [§3.2, Theorem 3.3] §3.2, Theorem 3.3: the uniqueness statement for the optimizing vector field dot T relies on the injectivity of the linear functional induced by the response; the proof sketch invokes the non-degeneracy condition but does not explicitly verify that this condition is open and dense in the space of observation functions c, which would strengthen the applicability claim.
  2. [§5.1, Eq. (5.4)] §5.1, Eq. (5.4): the error bound for the truncated Fourier scheme is stated in the anisotropic norm, yet the constant in the tail estimate appears to depend on the choice of the cutoff frequency N; a uniform bound independent of N (or an explicit dependence) is needed to confirm the convergence rate asserted in the theorem.
minor comments (3)
  1. [§2] The notation for the perturbation vector field dot T is introduced in §2 but occasionally overlaps with the time derivative symbol; a brief clarification in the notation paragraph would avoid confusion.
  2. [Figure 2] Figure 2 caption does not specify the value of the cutoff parameter N used in the numerical approximation; adding this detail would improve reproducibility.
  3. [§2.1] Reference [12] (Gouëzel-Liverani) is cited for the Banach-space setup, but the precise range of the anisotropy parameters (p,q) employed in the present work is not restated; repeating the interval would help readers.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful reading and constructive comments on our manuscript. We address each major comment below and have incorporated revisions to strengthen the presentation.

read point-by-point responses

  1. Referee: [§3.2, Theorem 3.3] §3.2, Theorem 3.3: the uniqueness statement for the optimizing vector field dot T relies on the injectivity of the linear functional induced by the response; the proof sketch invokes the non-degeneracy condition but does not explicitly verify that this condition is open and dense in the space of observation functions c, which would strengthen the applicability claim.

    Authors: We agree that an explicit verification of openness and density would strengthen the applicability claim. The non-degeneracy condition means that the linear response functional is nonzero. Since this functional is continuous on the space of smooth observation functions equipped with the C^∞ topology, its kernel is a closed hyperplane of codimension one. Consequently, the set of non-degenerate c is open and dense. We will add a short remark immediately after Theorem 3.3 stating and proving this fact. revision: yes

  2. Referee: [§5.1, Eq. (5.4)] §5.1, Eq. (5.4): the error bound for the truncated Fourier scheme is stated in the anisotropic norm, yet the constant in the tail estimate appears to depend on the choice of the cutoff frequency N; a uniform bound independent of N (or an explicit dependence) is needed to confirm the convergence rate asserted in the theorem.

    Authors: We thank the referee for this observation. Re-examining the proof, the constant in the tail estimate does depend on N via the parameters of the anisotropic norm. However, this dependence is explicit and of the form C(1 + log N) for a constant C independent of N. The error still tends to zero as N → ∞, confirming convergence. In the revised manuscript we update the statement of the theorem in §5.1 to display the explicit N-dependence in the error bound and clarify the corresponding estimate in the proof. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivation rests on established transfer-operator theory

full rationale

The paper invokes the standard linear response formula for SRB measures of Anosov diffeomorphisms (derivative of the transfer operator on Gouëzel-Liverani anisotropic Banach spaces) as an external, well-known input. The optimization problem over vector fields is then posed directly from this formula, uniqueness for non-degenerate response functions follows from injectivity of the resulting linear functional on the chosen perturbation space, and explicit Fourier coefficients are obtained by expanding the optimal vector field in a compatible basis and solving the ensuing linear system. The numerical scheme truncates this expansion with a convergence proof controlling the tail in the anisotropic norm. All steps are internally consistent with the cited functional-analytic framework; no quantity is defined in terms of itself, no fitted parameter is relabeled as a prediction, and no load-bearing premise reduces to a self-citation chain. The construction is therefore self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The work relies on standard background results in hyperbolic dynamics; no new free parameters or invented entities are introduced in the abstract.

axioms (1)
  • domain assumption Anosov diffeomorphisms possess linear response of their SRB measures with respect to infinitesimal perturbations of the map.
    Stated as well-known in the abstract and used as the foundation for the optimization problem.

pith-pipeline@v0.9.0 · 5664 in / 1237 out tokens · 44188 ms · 2026-05-22T18:41:51.190160+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Fixed-point approximation for self-consistent transfer operators with Newton's method

    math.DS 2026-05 unverdicted novelty 6.0

    Nonlinear Fourier-Fejér discretization yields convergent finite-dimensional approximations to fixed points of self-consistent transfer operators, with Newton's method delivering quadratic convergence.

Reference graph

Works this paper leans on

19 extracted references · 19 canonical work pages · cited by 1 Pith paper · 1 internal anchor

  1. [1]

    Antown, G

    F. Antown, G. Froyland, and S. Galatolo. Optimal linear response for Markov Hilbert- –Schmidt integral operators and stochastic dynamical systems.Journal of Nonlinear Science, 32:60, 2022

    work page 2022

  2. [2]

    V. Baladi. Linear response, or else. InProceedings of the International Congress of Mathematicians-Seoul 2014. Vol. III. Invited lectures, pages 525–545, 2014

    work page 2014

  3. [3]

    Brezis.Functional Analysis, Sobolev Spaces and Partial Differential Equations

    H. Brezis.Functional Analysis, Sobolev Spaces and Partial Differential Equations. Uni- versitext. Springer-Verlag New York, 1st edition, 2010

    work page 2010

  4. [4]

    Crimmins and G

    H. Crimmins and G. Froyland. Fourier approximation of the statistical properties of Anosov maps on tori.Nonlinearity, 33(11):6244, 2020

    work page 2020

  5. [5]

    Dragiˇ cevi´ c and G

    D. Dragiˇ cevi´ c and G. Froyland. Optimal linear responses for Markov chains and stochas- tically perturbed dynamical systems.Journal of Statistical Physics, 170:1051–1087, 2018. 32

    work page 2018

  6. [6]

    Program Generation, Optimization, and Platform Adaptation

    M. Frigo and S. G. Johnson. The design and implementation of FFTW3.Proceedings of the IEEE, 93(2):216–231, 2005. Special issue on “Program Generation, Optimization, and Platform Adaptation”

    work page 2005

  7. [7]

    Froyland and S

    G. Froyland and S. Galatolo. Optimal linear response for expanding circle maps.Non- linearity, 38(3):035001, 2025

    work page 2025

  8. [8]

    Froyland, P

    G. Froyland, P. Koltai, and M. Stahn. Computation and optimal perturbation of finite- time coherent sets for aperiodic flows without trajectory integration.SIAM Journal on Applied Dynamical Systems, 19(3):1659–1700, 2020

    work page 2020

  9. [9]

    Galatolo and A

    S. Galatolo and A. Ni. Optimal response for hyperbolic systems by the fast adjoint response method.arXiv preprint arXiv:2501.02395, 2025

    work page arXiv 2025

  10. [10]

    Galatolo and M

    S. Galatolo and M. Pollicott. Controlling the statistical properties of expanding maps. Nonlinearity, 30(7):2737–2751, 2017

    work page 2017

  11. [11]

    Ghil and V

    M. Ghil and V. Lucarini. The physics of climate variability and climate change.Rev. Mod. Phys., 92:035002, Jul 2020

    work page 2020

  12. [12]

    Gou¨ ezel and C

    S. Gou¨ ezel and C. Liverani. Banach spaces adapted to Anosov systems.Ergodic Theory and dynamical systems, 26(1):189–217, 2006

    work page 2006

  13. [13]

    Hairer and A

    M. Hairer and A. J. Majda. A simple framework to justify linear response theory. Nonlinearity, 23(4):909, 2010

    work page 2010

  14. [14]

    Keller and C

    G. Keller and C. Liverani. Stability of the spectrum for transfer operators.Annali della Scuola Normale Superiore di Pisa-Classe di Scienze, 28(1):141–152, 1999

    work page 1999

  15. [15]

    W. M. Lai, D. Rubin, and E. Krempl.Introduction to continuum mechanics. Butterworth-Heinemann, 3rd edition, 1996

    work page 1996

  16. [16]

    M. Porte. Linear response for Dirac observables of Anosov diffeomorphisms.Discrete Contin. Dyn. Syst., 39(4):1799–1819, 2019

    work page 2019

  17. [17]

    Forward-Mode Automatic Differentiation in Julia

    J. Revels, M. Lubin, and T. Papamarkou. Forward-mode automatic differentiation in Julia.arXiv:1607.07892 [cs.MS], 2016

    work page internal anchor Pith review Pith/arXiv arXiv 2016

  18. [18]

    D. Ruelle. Differentiation of SRB states.Communications in Mathematical Physics, 187:15, 1997

    work page 1997

  19. [19]

    Sebastien and C

    G. Sebastien and C. Liverani. Compact locally maximal hyperbolic sets for smooth maps: fine statistical properties.Journal of Differential Geometry, 79:433–477, 2006. 33

    work page 2006