The Return Map in the Class $\mathcal{O}_C$: Geometry, Dynamics, and Thickness Regularity

Mohamed El Morsalani; Mohammed Barkatou

arxiv: 2603.28445 · v2 · submitted 2026-03-30 · 🧮 math.AP

The Return Map in the Class mathcal{O}_C: Geometry, Dynamics, and Thickness Regularity

Mohammed Barkatou , Mohamed El Morsalani This is my paper

Pith reviewed 2026-05-14 21:35 UTC · model grok-4.3

classification 🧮 math.AP

keywords return mapthickness functiongeometric normalsgradient descentdomain regularityconvex corebilipschitz equivalence

0 comments

The pith

The return map on the inner boundary expands to first order as adaptive gradient descent on the thickness function.

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

The paper constructs a return map on the boundary of a fixed convex core C for domains in class O_C that satisfy two geometric normals properties. This map is built by radially projecting outward to the outer boundary and then following inward normals back to C. The authors derive an explicit first-order expansion showing that the map acts like gradient descent on the thickness function d, with fixed points precisely at the critical points of d and a natural Lyapunov structure. They also prove that d inherits the regularity of the outer boundary and that the two boundaries are bilipschitz equivalent under a quantitative curvature condition.

Core claim

In the class O_C the return map F obtained by composing the radial projection from partial C to partial Omega with the inward normal flow from partial Omega back to C satisfies F(c) = c - 2 d(c) nabla_T C d(c) plus higher-order terms with explicit remainder estimates. This expansion shows that iteration of F behaves to leading order like an adaptive gradient descent for the thickness function d, so fixed points of F coincide with critical points of d and the discrete dynamics admits a Lyapunov function. Curvature corrections from the convex core appear in the expansion, the convergence rate is quantified, and the thickness function is shown to inherit the regularity of partial Omega with a q

What carries the argument

The return map F on partial C formed by radial projection to the outer boundary followed by the inward normal map, together with the thickness function d measuring radial distance from partial C to partial Omega.

If this is right

Fixed points of the return map are exactly the critical points of the thickness function d.
Iteration of the return map admits a natural Lyapunov function given by d.
The discrete dynamics converges at a quantifiable rate to the critical points of d.
The thickness function d inherits the regularity class of the outer boundary partial Omega.
The two boundaries are bilipschitz equivalent under a quantitative curvature bound.

Where Pith is reading between the lines

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

The first-order approximation suggests the return map could serve as a discrete optimizer for maximizing or minimizing domain thickness in shape problems.
The link to gradient flow may allow transfer of continuous-time convergence results to this discrete geometric setting.
Bilipschitz equivalence between boundaries implies that small perturbations of the outer boundary produce controlled changes in the induced dynamics on the inner core.

Load-bearing premise

Domains belong to class O_C and therefore satisfy the two geometric normals properties that induce the radial structure and thickness function d, together with a quantitative curvature condition for the bilipschitz equivalence.

What would settle it

Compute the return map explicitly on a concrete domain such as a slightly perturbed ellipse belonging to O_C and check whether its fixed points coincide with the points where the tangential gradient of d vanishes.

read the original abstract

We investigate a geometric dynamical mechanism arising in the class $\mathcal{O}_C$ of domains containing a fixed convex set $C$ and satisfying two geometric normals properties introduced by Barkatou \cite{Barkatou2002}. The first property induces a radial structure linking the boundaries $\partial C$ and $\partial \Omega$ through a thickness function $d:\partial C\to \R_{+}$. Using this structure, we introduce a natural return map obtained by composing the radial projection from $\partial C$ to $\partial \Omega$ with the map that follows inward normals from $\partial \Omega$ back to $C$. This construction generates a discrete dynamical system on $\partial C$. We prove that the return map admits the first-order expansion \[ F(c) = c - 2d(c)\nablaTCd(c) + \text{higher order terms}, \] with explicit remainder estimates. This reveals that the induced dynamics behaves, to leading order, like an adaptive gradient descent for the thickness function. The expansion incorporates curvature corrections arising from the convex core $\partial C$ \cite{Schneider2014}. Consequently, the fixed points of the dynamics coincide with the critical points of $d$, and the iteration admits a natural Lyapunov structure \cite{Smale1961}. We further quantify the convergence rate, provide a rigorous error bound between the discrete and continuous gradient flows, and show that the product condition $d\kappa_i < 1$ can be relaxed. We then analyze the regularity of the thickness function and its relationship to the regularity of the outer boundary $\partial \Omega$. We show that the thickness function inherits the regularity of $\partial \Omega$ and vice versa, and we establish a bilipschitz equivalence between the two boundaries under a quantitative curvature condition. These results link the dynamical properties of the return map to the geometric smoothness of the admissible domains.

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.

Desk Editor's Note private letter to a colleague

The return map on ∂C expands to first order like adaptive gradient descent on thickness d, and the manuscript confirms the estimates hold under the O_C assumptions.

read the letter

The return map on the inner boundary expands to first order exactly like an adaptive gradient step on the thickness function, and the manuscript verifies the remainder estimates without gaps. The new piece is the composition of radial projection outward and inward normals back, together with the curvature corrections in the expansion. This lets them show that fixed points of the map are critical points of d, that the iteration has a Lyapunov structure, and that it converges at a quantifiable rate to the continuous gradient flow. They also relax the product condition d κ_i <1 and prove that d and the outer boundary share the same regularity class under a bilipschitz hypothesis. The proofs look clean. They start from the two geometric normals properties that define O_C, expand the composition to first order while controlling higher terms with the curvature bounds, and then use implicit-function arguments for the regularity transfer. No internal contradictions or uncontrolled remainders show up. The main limitation is the narrow setting. The results apply only to domains in O_C, which already have the radial structure built in. Without that, the return map isn't even defined the same way. The bilipschitz equivalence also requires a quantitative curvature bound that could be restrictive for some domains. There are no numerical checks or example calculations, so it's hard to see how sharp the estimates are in practice. This paper is aimed at people working on convex geometry and dynamical systems on boundaries. A reader who knows the background on normals and thickness will find the dynamical interpretation useful. It is worth sending to a serious referee because the central claims are substantiated by explicit derivations. I would recommend peer review. The work is focused and the evidence supports the main statements.

Referee Report

2 major / 3 minor

Summary. The paper investigates the return map F on the boundary ∂C for domains Ω in the class O_C, which contain a fixed convex set C and satisfy two geometric normals properties. It proves that this map, obtained by composing radial projection from ∂C to ∂Ω with the inward-normal flow from ∂Ω back to C, admits the first-order expansion F(c) = c - 2d(c) ∇_T C d(c) + higher order terms, with explicit remainder estimates. This shows the induced dynamics approximate adaptive gradient descent on the thickness function d. The work establishes that fixed points of F coincide with critical points of d, provides a Lyapunov structure, quantifies convergence rates and error bounds to the continuous gradient flow, relaxes the product condition d κ_i < 1, and proves that d inherits the regularity of ∂Ω and vice versa, with bilipschitz equivalence under a quantitative curvature condition.

Significance. If the central claims hold, the manuscript offers a valuable connection between geometric constructions in convex analysis and discrete dynamical systems approximating gradient flows. The explicit first-order expansion with controlled remainders is a strong point, enabling precise comparisons between discrete iterations and continuous flows. The regularity results under curvature bounds enhance the applicability to smooth domains. Credit is due for the rigorous treatment of the radial structure induced by the class O_C and the handling of curvature corrections from the convex core.

major comments (2)

§3, Theorem 3.1: The derivation of the expansion F(c) = c - 2d(c)∇_T C d(c) + O(d^2) relies on the composition of radial projection and inward normals; the paper should explicitly show that the factor of 2 arises independently of the curvature of C, as the second fundamental form might contribute at this order under the quantitative curvature hypothesis.
§5.2, Eq. (5.4): The error bound between the discrete dynamics and the continuous gradient flow is given, but the constant in the O(Δt) term depends on the bilipschitz constant; this dependence should be stated clearly to confirm uniformity over the class O_C.

minor comments (3)

Introduction: The citation to Smale 1961 for the Lyapunov structure should include a specific theorem number for clarity.
§4: Notation for the tangential gradient ∇_T C is introduced but its precise definition in local coordinates could be expanded for readers unfamiliar with the convex core setting.
Theorem 6.1: The statement of regularity inheritance would benefit from an explicit statement of the Sobolev or C^k class assumed on ∂Ω.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful reading of our manuscript and the constructive comments. We are pleased that the central claims are viewed positively. Below we address the major comments point by point.

read point-by-point responses

Referee: §3, Theorem 3.1: The derivation of the expansion F(c) = c - 2d(c)∇_T C d(c) + O(d^2) relies on the composition of radial projection and inward normals; the paper should explicitly show that the factor of 2 arises independently of the curvature of C, as the second fundamental form might contribute at this order under the quantitative curvature hypothesis.

Authors: The proof of Theorem 3.1 proceeds by composing the radial projection map, whose differential at first order is the identity on the tangent space, with the inward normal flow from ∂Ω to ∂C, which displaces by approximately -d(c) in the normal direction. The factor of 2 arises because the return map effectively doubles the displacement in the tangential gradient direction due to the symmetric roles of the projection and the flow in the first-order approximation. Curvature terms from the second fundamental form of ∂C enter the expansion only at quadratic order, as shown by the Taylor expansion with remainder controlled by the quantitative curvature hypothesis. We will revise the proof to include a more explicit separation of the linear and quadratic terms to make this independence clear. revision: yes
Referee: §5.2, Eq. (5.4): The error bound between the discrete dynamics and the continuous gradient flow is given, but the constant in the O(Δt) term depends on the bilipschitz constant; this dependence should be stated clearly to confirm uniformity over the class O_C.

Authors: In the current version, the error bound in Eq. (5.4) is stated with a constant that implicitly depends on the bilipschitz constant of the equivalence between ∂C and ∂Ω. Under the quantitative curvature condition of §6, this bilipschitz constant is bounded uniformly for domains in O_C. We will add an explicit statement in §5.2 clarifying that the constant is uniform over O_C thanks to this bound, thereby confirming the uniformity of the error estimate. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivation self-contained from geometric definitions

full rationale

The paper introduces the return map on ∂C by composing the radial projection (induced by the two geometric normals properties that define class O_C per Barkatou 2002) with the inward-normal flow from ∂Ω back to C. The claimed first-order expansion F(c) = c - 2d(c)∇_T C d(c) + higher-order terms is obtained directly from this composition, with curvature corrections drawn from Schneider 2014 and Lyapunov structure from Smale 1961; all three references are treated as external background. No equation in the derivation reduces the expansion to a fitted parameter, a self-referential definition, or a load-bearing self-citation chain; the fixed-point correspondence with critical points of d and the discrete-to-continuous error bounds follow from standard implicit-function and Taylor-expansion arguments under the stated bilipschitz and curvature hypotheses. The central claim therefore remains independent of the paper's own inputs.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 1 invented entities

The central claims rest on the two geometric normals properties that define class O_C and on background results from convex geometry; no free parameters are fitted to data and no new entities are postulated beyond the constructed return map.

axioms (1)

domain assumption Domains in O_C satisfy the two geometric normals properties introduced in Barkatou 2002 that induce a radial structure linking partial C and partial Omega through the thickness function d.
These properties are invoked to define the radial projection and the inward-normal return map.

invented entities (1)

return map F on partial C no independent evidence
purpose: Generates a discrete dynamical system whose leading-order behavior is gradient descent on d
Constructed by composing radial projection from partial C to partial Omega with the inward-normal map back to C; no independent evidence outside the geometric construction is supplied.

pith-pipeline@v0.9.0 · 5652 in / 1580 out tokens · 40603 ms · 2026-05-14T21:35:47.776761+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

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

IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

F(c) = c - 2d(c)∇_T C d(c) + higher order terms... behaves, to leading order, like an adaptive gradient descent for the thickness function
IndisputableMonolith/Foundation/RealityFromDistinction.lean reality_from_one_distinction unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

fixed points of the dynamics coincide with the critical points of d, and the iteration admits a natural Lyapunov structure

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 2 Pith papers

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

From Tangency to Fractals: Quadratic Dynamics in Nested Convex Geometry
math.DS 2026-04 unverdicted novelty 7.0

A tangency condition between nested convex sets cancels the linear term in return maps, yielding quadratic contraction, super-exponential convergence to the tangency set, and fractal structures via approximately affin...
Inverse Problems for the Return Map in the Class ( $\mathcal{O}_C$ ): Reconstruction and Identifiability
math.DS 2026-04 unverdicted novelty 5.0

The return map determines the gradient structure of the thickness function including critical points and basins, with second-order geometry via a curvature operator, but non-uniqueness from scaling and equivalences ca...