Pointers

XII Bienal SBM · UFRN · 3–7 August 2026

Seven plenaries on this programme touch the dm³ work directly. This page says where, and what I would actually ask. It is a list of questions, not of claims.

The system, in one paragraph

dm³ is a toy ODE on the contact manifold M = ℝ²₊ × ℝ: ṙ = r(1−r²) + 2(r−1)e⁻ᶻ, θ̇ = 1, ż = r² − 2(r−1)²e⁻ᶻ. It has an attracting limit cycle at r = 1. A Grönwall estimate gives a symmetric ball of guaranteed convergence, |r−1| < 1/3. High-precision DOP853 integration says the real inner boundary is at r★ = 0.77594058, not 2/3. The basin is asymmetric, and the symmetric estimate does not merely lose sharpness — it misclassifies orbits. That gap is the whole pedagogical content of the minicurso.

Contact geometry and Reeb dynamics

Pedro Salomão · SUSTech
Friday 7 Aug · 08:00 · Plenária 13 — Os caminhos invisíveis da Lua: do problema de Hill à geometria moderna

Hill's problem to modern geometry is global surfaces of section and Reeb dynamics on contact manifolds. dm³ is a dissipative cousin: the same ambient object, without the Hamiltonian.

Ask: the inner basin boundary is a global statement about a limit cycle, obtained numerically. Do global surface-of-section methods give it as a theorem — and does dissipation break the machinery, or only complicate it?

Planar vector fields · the shortest route to a proof

Regilene Delazari dos Santos Oliveira · USP
Tuesday 4 Aug · 10:30 · Plenária 6 — On topological equivalence of planar vector fields induced by the Newton polyhedron

Because θ̇ = 1 decouples, dm³ reduces to a planar system in (r, z). That puts it inside the class this talk is about.

Ask: can the Newton polyhedron of the (r,z) field bound the inner basin analytically? A polyhedron argument would replace a numerical certificate with a proof — the single most valuable thing anyone could hand this project.

Singularities

Débora Lopes · UFS
Friday 7 Aug · 09:00 · Plenária 14 — O que há de especial nas singularidades?

Volume I classifies the dm³ bifurcations against Whitney A₁–A₃, and the A₁ node is the bridge file between DM3-lab and AXLE across seven physical instances.

Ask: is Whitney the right classification for a dissipative contact system, or is it being borrowed from a setting where it means something stronger?

Continued fractions and the n-bonacci constants

Carlos Gustavo Moreira (Gugu) · IMPA
Wednesday 5 Aug · 09:00 · Plenária 8 — Frações contínuas e aproximações de números reais por números racionais

The DNLS chain work turns on the tribonacci constant η ≈ 1.8393 and its tetrabonacci successor, with a criticality threshold observed at that value.

Ask: is the criticality threshold an approximation-theoretic statement in disguise? If the localisation transition tracks how badly η is approximable by rationals, the numerics have been measuring a number-theoretic quantity all along.

Geometric flows and the curvature threshold

Ronaldo Freire de Lima · UFRN
Thursday 6 Aug · 09:00 · Plenária 11 — Fluxos Geométricos

The K operator in the dm³ chain is a curvature operator with a verified threshold κ* ≤ √(7/9) ≈ 0.882. The framework treats curvature as the thing that removes options rather than the thing that drives motion.

Ask: does the threshold have a flow interpretation — is κ* a bound that a geometric flow would reach, or an artefact of the normal form?

Cyclotomy and the sixfold constants

Rodrigo Gondim · UFRPE
Tuesday 4 Aug · 08:00 · Plenária 4 — Ciclotomia

The G6 Crystal is a hexagonal form with aspect ratio 66 = 33·τ, and the verified constants include g₆₄ = 2⁶ = 64 and T* = 2π. Sixfold symmetry throughout.

Ask: is g₃₃ = 33 a cyclotomic object, or numerology that has survived because nobody with the right training has looked at it? An honest answer either way is worth more than the constant.

Teaching · why CO144 exists

Victor Giraldo · UFRJ
Friday 7 Aug · 13:00 (PROFMAT) and 16:30 (Mesa redonda — Formação de Formadores)

CO144 argues that English for mathematical researchers can be taught through the mathematics itself, with science entering from day 21 of instruction rather than after fluency.

Ask: where does that sit against saberes de matemática para o ensino? The claim needs to survive this framework, not avoid it.

One honest gap

AXLE carries over 100 Lean 4 theorems with zero sorry in the verified core. The basin boundary is not among them. The Lean statement currently proves only that some ρ★ exists in the interval (2/3, 1) — which 0.77594058 satisfies, and so does 0.773, and so does 0.8. The value itself rests on DOP853 integration, not on a proof. It is a certificate, not a theorem, and it is named as such in the file. Closing that gap is the open problem this exhibition is advertising.

Elsewhere in the wider project