{
 "artifacts": null,
 "category": "dynamics",
 "description": "Replace an unconstrained third-order stationarity update by a Chebyshev root step on a one-dimensional restriction of the neural loss, while monitoring the images of the update map's free critical points. Shrink the trust radius or damping parameter whenever the maximum critical-value gain exceeds one, because the paper's mechanism predicts that critical-orbit expansion marks loss of a safe attracting basin.",
 "formulas_latex": [
  "$$C_p(x)=x-(1+L_p(x)/2)p(x)/p'(x),\\quad L_p(x)=p(x)p''(x)/(p'(x))^2.$$",
  "$$c^*=C_n(c)=-D_n c,\\quad D_n=((n-1)^2(2n+1))/(27(n+1)^2).$$",
  "$$Gamma=max_{c in Crit_free(C_hat)} |C_hat(c)-t_*|/|c-t_*|,\\quad accept\\ if\\ Gamma\u003c=gamma\u003c1.$$",
  "$$t_{k+1}=t_k-alpha(1+hatL(t_k)/2)hatp(t_k)/hatp'(t_k),\\quad theta_{k+1}=theta_k+t_{k+1}u.$$"
 ],
 "id": 3093,
 "implementation": "Integrate this as an optional inner update for a neural optimizer rather than replacing minibatch SGD everywhere. At parameter vector theta, choose a normalized direction u=-P g_hat/||P g_hat||, where g_hat is a minibatch gradient and P is a diagonal Adam-style preconditioner. Probe the directional stationarity function p(t)=u^T grad L(theta+t u) at five to nine small values of t, using the same minibatch or common random numbers. Fit a degree-3 or degree-5 polynomial p_hat(t), differentiate it analytically, and apply the Chebyshev update. Find the real roots of the derivative C_hat'(t) with a polynomial root solver; discard poles, points outside the trust interval [-r,r], and critical points close to the selected root t_*. Compute Gamma from the fitted polynomial. If Gamma exceeds 0.9, halve r and set alpha to 0.5 alpha. If Gamma is below 0.5, increase r by 10 percent up to a cap. Reject any update that increases a fresh validation minibatch loss. The exact Chebyshev map and the contraction-expansion mechanism come from the paper; polynomial fitting, root selection, and the threshold are empirical adaptations. First test on a two-layer MLP fitting a one-dimensional polynomial regression dataset, comparing SGD, Adam, and undamped Chebyshev updates. Sweep trust radii and record divergence. The quantitative prediction is a transition near Gamma=1: Gamma\u003c1 should produce bounded directional iterates and cubic local stationarity convergence, whereas Gamma\u003e1 should produce more basin exits or exploding steps. On the exact synthetic family p_n, the implementation should recover the transition between n=16 and n=17 when D_n crosses one.",
 "math_summary": "For a scalar stationarity equation p(x)=0, Chebyshev's method is C_p(x)=x-(1/2 L_p(x)+1)p(x)/p'(x), where L_p(x)=p(x)p''(x)/(p'(x))^2. For p_n(z)=z(z^n-1), every free critical point c of C_n has critical value c*=C_n(c)=-D_n c, with D_n=(n-1)^2(2n+1)/(27(n+1)^2). Therefore |c*|/|c|=D_n is below one for n\u003c=16 and above one for n\u003e=17. In the neural adaptation, p(t)=u^T grad_theta L(theta+t u) is a directional stationarity function, u is a normalized update direction, and p'(t) and p''(t) are directional second and third derivatives. Fit a local polynomial p_hat(t) and compute its Chebyshev map C_hat. Define Gamma as the largest ratio of the distance from a free critical point's image to the selected local root over the original distance. Use Gamma below one as an empirical contraction certificate.",
 "math_tags": [
  "dynamical-systems",
  "bifurcations",
  "optimization",
  "linear-algebra"
 ],
 "ml_areas": [
  "optimizer",
  "training-dynamics",
  "loss"
 ],
 "paper": {
  "arxiv_id": "2609.02884",
  "arxiv_url": "https://arxiv.org/abs/2609.02884",
  "summary_what_math_gives_to_ml": "The paper provides a sharp dynamical criterion for global behavior of Chebyshev's third-order root iteration on the rotationally symmetric family p_n(z)=z(z^n-1). Free critical points map radially by a constant factor D_n=(n-1)^2(2n+1)/(27(n+1)^2), producing a transition at n=17: critical values contract for n\u003c=16 and expand for n\u003e=17. This mechanism can be transferred to higher-order neural optimizers by treating stationarity along a training direction as a local polynomial root problem and using the maximum critical-orbit gain as a trust-region or step-size controller. The key experiment is to test whether the predicted critical-gain boundary separates stable third-order updates from exploding or basin-leaving updates.",
  "title": "Chebyshev's method applied to polynomials with rotational symmetry",
  "year": "2026"
 },
 "ratings": {
  "difficulty": 6,
  "novelty": 8,
  "usefulness": 6
 },
 "solves": [
  "stability",
  "speedup"
 ],
 "title": "Critical-Orbit Trust Region for Chebyshev Optimizers",
 "url": "https://synthcore.org/idea/3093/critical-orbit-trust-region-for-chebyshev-optimizers",
 "verification": {
  "peer_reviewed": false,
  "status": "unverified",
  "status_label": "Unverified",
  "verdict_source": "deterministic test code (paired-seed permutation statistics)",
  "verification_axes": {
   "benchmark_mechanism": {
    "confirmed": null,
    "tested": false
   },
   "practical_benchmark": {
    "beats_baseline": null,
    "tested": false
   },
   "toy_mechanism_gate": {
    "confirmed": null,
    "tested": false
   }
  }
 }
}