pith. sign in

arxiv: 2606.25440 · v1 · pith:7UH2A2HYnew · submitted 2026-06-24 · 🧮 math.CA · math.DS· math.PR

Dragon curves in Littlewood roots

Marcus Michelen , Oren Yakir This is my paper

Pith reviewed 2026-06-25 20:03 UTC · model grok-4.3

classification 🧮 math.CA math.DSmath.PR
keywords Littlewood polynomialsdragon curvesiterated function systemspolynomial rootsfractal attractorscomplex analysisroot distributions
0
0 comments X

The pith

Roots of all Littlewood polynomials of large degree form dragon curve fractals away from the unit circle.

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

The paper establishes that the collection of roots from every Littlewood polynomial of large degree produces fractal patterns away from the unit circle. These patterns match the attractor of an iterated function system, and the same structure appears in a random model of the polynomials. A sympathetic reader would care because the work converts a visual numerical curiosity into a proven typical property of the root distributions.

Core claim

When the roots of all Littlewood polynomials of a given large degree are plotted together, they form striking fractal structures away from the unit circle that match the attractor of a certain iterated function system, with analysis of a random variant showing that such fractal behavior is typical.

What carries the argument

The attractor of an iterated function system that captures the limiting distribution of roots for both deterministic and random Littlewood polynomials.

If this is right

  • The fractal structures appear consistently for both the standard Littlewood polynomials and their random variants.
  • The majority of roots remain near the unit circle while the off-circle roots follow the fractal attractor.
  • The phenomenon receives a rigorous proof rather than remaining a numerical observation.

Where Pith is reading between the lines

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

  • Similar iterated function system models may describe root distributions for other coefficient-restricted polynomial families.
  • The typicality result could guide statistical sampling methods when studying large ensembles of random polynomials.
  • Higher-degree computations could be checked directly against the attractor to measure convergence rate.

Load-bearing premise

The observed numerical fractals are generated by an iterated function system whose attractor accurately captures the limiting distribution of roots for both the deterministic and random models of Littlewood polynomials.

What would settle it

A computation of the empirical root distribution from all Littlewood polynomials of sufficiently high degree that fails to converge to the attractor of the iterated function system would falsify the explanation.

Figures

Figures reproduced from arXiv: 2606.25440 by Marcus Michelen, Oren Yakir.

Figure 1
Figure 1. Figure 1: Heat map for all the roots of polynomials with ±1 coefficients of degree 26. A Littlewood polynomial is a polynomial with coefficients in {−1, +1}. We denote by Ln the set of all Littlewood polynomials of degree n ≥ 1. Let D = {z ∈ C : |z| < 1} denote the open unit disk. In this paper, we study the (microscopic) structure of the set of all possible roots of Littlewood polynomials of large degree in the dis… view at source ↗
Figure 2
Figure 2. Figure 2: Top row: Dragon curves for three different choices of the parameter α. Bottom row: zoomed-in view near the same points (in red) from [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Top row: roots of the degree-12 polynomials p1, p2, p3 (red), together with the roots of all degree-24 Littlewood polynomials extending them (blue). Bottom row: zoomed-in view near the corresponding roots α1, α2, α3. Note that the roots resemble the dragons from [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
read the original abstract

A Littlewood polynomial is a polynomial whose coefficients lie in $\{- 1, +1\}$. While the majority of roots of a Littlewood polynomial of large degree are near the unit circle, numerical experiments suggest that when plotting the roots of \emph{all} Littlewood polynomials of a given large degree, striking fractal structures appear away from the unit circle. These fractals resemble the attractor of a certain iterated function system and are known as \emph{dragon curves}. In this note, we provide a rigorous explanation of this phenomenon, along with an analysis of a random variant, saying that such fractal behavior is typical.

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 / 2 minor

Summary. The manuscript claims that plotting the roots of all Littlewood polynomials (coefficients in {-1,+1}) of large fixed degree produces fractal structures away from the unit circle that resemble dragon curves; these are identified with the attractor of a specific iterated function system (IFS). It supplies a rigorous explanation of the phenomenon for the deterministic enumeration of all such polynomials and for a random variant, concluding that fractal behavior is typical.

Significance. If the claimed rigorous identification of the IFS attractor with the support and limiting distribution of the roots holds, the work would furnish a precise geometric explanation for the observed numerical fractals in Littlewood root plots and would establish that such structure persists under randomization. The explicit treatment of both the deterministic and random models is a positive feature.

major comments (2)
  1. [Section constructing the IFS maps and stating the main theorem] The central claim requires that the IFS attractor is not merely a visual or set-theoretic container but the actual support (and ideally the weak limit) of the empirical root measure. The provided argument appears to establish containment of limit points inside the attractor via the recursive sign choices, but does not demonstrate that every point of the attractor is realized as a limit point of actual roots or that the empirical measures converge weakly to the IFS invariant measure.
  2. [Analysis of the random variant] For the random variant, the analysis shows that the same IFS governs the typical behavior, yet it is unclear whether the proof controls the discrepancy between the random sign model and the deterministic enumeration sufficiently to transfer the limiting distribution statement without additional error estimates.
minor comments (2)
  1. The notation for the individual IFS contractions could be introduced with an explicit list of the four maps and their fixed-point equations to aid readers.
  2. A brief remark on how the numerical plots were generated (e.g., degree range and root-finding method) would improve reproducibility of the motivating figures.

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 will revise the paper accordingly to strengthen the arguments.

read point-by-point responses

  1. Referee: [Section constructing the IFS maps and stating the main theorem] The central claim requires that the IFS attractor is not merely a visual or set-theoretic container but the actual support (and ideally the weak limit) of the empirical root measure. The provided argument appears to establish containment of limit points inside the attractor via the recursive sign choices, but does not demonstrate that every point of the attractor is realized as a limit point of actual roots or that the empirical measures converge weakly to the IFS invariant measure.

    Authors: We agree that a more explicit demonstration is needed to confirm the attractor coincides with the support of the limiting root set. In the revision we will insert a lemma constructing, for any point in the attractor, a sequence of sign patterns whose associated roots converge to it by iterating the inverse branches of the IFS maps. We will also add a short argument establishing weak convergence of the empirical root measures to the unique IFS-invariant probability measure, using the contractive property and uniqueness of the invariant measure. revision: yes

  2. Referee: [Analysis of the random variant] For the random variant, the analysis shows that the same IFS governs the typical behavior, yet it is unclear whether the proof controls the discrepancy between the random sign model and the deterministic enumeration sufficiently to transfer the limiting distribution statement without additional error estimates.

    Authors: The random-model analysis establishes that the same IFS governs the roots with high probability. To close the gap with the deterministic enumeration we will add explicit discrepancy bounds (e.g., via concentration or Wasserstein-distance estimates) showing that the random empirical measures remain close to their deterministic counterparts, thereby transferring the weak-convergence statement to the random setting. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivation is self-contained from polynomial recursion

full rationale

The paper derives an iterated function system directly from the recursive sign-choice structure inherent to Littlewood polynomials (coefficients in {-1,+1}), then rigorously analyzes the attractor for both the deterministic enumeration and a random variant. This construction is independent of the target limiting distribution; the IFS maps are defined from the same algebraic recursion that generates the polynomials, and the claim that fractal behavior is typical follows from explicit analysis of the attractor rather than any fitted parameter, self-citation chain, or renaming of an input. No load-bearing step reduces by construction to the observed roots or their empirical measure. The derivation therefore stands as a genuine mathematical explanation rather than a tautology.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract alone supplies no information on free parameters, background axioms, or new entities; all such items are unknown.

pith-pipeline@v0.9.1-grok · 5621 in / 902 out tokens · 21013 ms · 2026-06-25T20:03:36.806894+00:00 · methodology

discussion (0)

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

Reference graph

Works this paper leans on

20 extracted references

  1. [1]

    Ahlfors.Complex analysis—an introduction to the theory of analytic functions of one complex variable

    L. Ahlfors.Complex analysis—an introduction to the theory of analytic functions of one complex variable. AMS Chelsea Publishing, 2021

    2021

  2. [2]

    J. C. Baez, J. D. Christensen, and S. Derbyshire. The beauty of roots.Notices Amer. Math. Soc., 70(9):1495–1497, 2023

    2023

  3. [3]

    C. Bandt. On the Mandelbrot set for pairs of linear maps.Nonlinearity, 15(4):1127–1147, 2002

    2002

  4. [4]

    M. F. Barnsley and A. N. Harrington. A Mandelbrot set for pairs of linear maps.Phys. D, 15(3):421–432, 1985

    1985

  5. [5]

    Beer.Topologies on closed and closed convex sets

    G. Beer.Topologies on closed and closed convex sets. Kluwer Academic Publishers Group, Dordrecht, 1993

    1993

  6. [6]

    C. J. Bishop and Y. Peres.Fractals in probability and analysis. Cambridge University Press, 2017

    2017

  7. [7]

    Borwein.Computational excursions in analysis and number theory, volume 10

    P. Borwein.Computational excursions in analysis and number theory, volume 10. Springer-Verlag, New York, 2002

    2002

  8. [8]

    Borwein and C

    P. Borwein and C. Pinner. Polynomials with 0, +1, -1 coefficients and a root close to a given point.Canadian Journal of Mathematics, 49(5):887–915, 1997

    1997

  9. [9]

    T. Bousch. Paires de similitudes. Unpublished manuscript.https://www.imo.universite-paris-saclay.fr/ ~thierry. bousch/preprints/. 1988

    1988

  10. [10]

    Calegari, S

    D. Calegari, S. Koch, and A. Walker. Roots, Schottky semigroups, and a proof of Bandt’s conjecture.Ergodic Theory Dynam. Systems, 37(8):2487–2555, 2017

    2017

  11. [11]

    P. Erd˝ os. On a lemma of Littlewood and Offord.Bull. Amer. Math. Soc., 51:898–902, 1945

    1945

  12. [12]

    Erd˝ os and P

    P. Erd˝ os and P. Tur´ an. On the distribution of roots of polynomials.Ann. of Math. (2), 51:105–119, 1950

    1950

  13. [13]

    D. Hokken. Topology of zero sets of polynomials with square discriminant.C. R. Math. Acad. Sci. Paris, 364:101–106, 2026

    2026

  14. [14]

    S. V. Konyagin and W. Schlag. Lower bounds for the absolute value of random polynomials on a neighborhood of the unit circle.Trans. Amer. Math. Soc., 351(12):4963–4980, 1999

    1999

  15. [15]

    J. E. Littlewood. On polynomials Pn ±zm, Pn eαmizm,z=e θi.J. London Math. Soc., 41:367–376, 1966

    1966

  16. [16]

    J. E. Littlewood and A. C. Offord. On the number of real roots of a random algebraic equation. III.Rec. Math. [Mat. Sbornik] N.S., 12/54:277–286, 1943

    1943

  17. [17]

    Michelen and O

    M. Michelen and O. Yakir. Limit law for root separation in random polynomials.Adv. Math., 496:Paper No. 110990, 2026

    2026

  18. [18]

    A. M. Odlyzko and B. Poonen. Zeros of polynomials with 0,1 coefficients.Enseign. Math. (2), 39(3-4):317–348, 1993

    1993

  19. [19]

    Shmerkin and B

    P. Shmerkin and B. Solomyak. Absolute continuity of complex Bernoulli convolutions.Math. Proc. Cambridge Philos. Soc., 161(3):435–453, 2016

    2016

  20. [20]

    Solomyak and H

    B. Solomyak and H. Xu. On the ‘Mandelbrot set’ for a pair of linear maps and complex Bernoulli convolutions.Nonlinearity, 16(5):1733–1749, 2003. Department of Mathematics, Northwestern University Email address:michelen@northwestern.edu Department of Mathematics, Massachusetts Institute of Technology Email address:oren.yakir@gmail.com

    2003