{
 "artifacts": null,
 "category": "dynamics",
 "description": "Run an ensemble of noisy optimization trajectories and regard trajectories that return to the same loss basin as competing dynamical phases. Estimate a complex return generating function from their path costs; a near-zero of this function signals cancellation between trajectory families and predicts an abrupt change in basin occupancy. Use the signal to reduce learning rate or optimizer noise near a transition, or increase noise when one phase dominates too early.",
 "formulas_latex": [
  "$$Z_q(t)=\\langle\\psi_0|e^{-iHt/\\hbar}|\\psi_0\\rangle.$$",
  "$$Z_c(T;s)=\\int_{x(0),x(T)\\in B_0}\\mathcal{D}x\\;\\exp\\left[-\\frac{\\mathcal{A}[x]}{D}-s\\mathcal{R}[x]\\right].$$",
  "$$Z_c(T;s)\\approx C_1(s)e^{-\\Phi_1(s)/D}+C_2(s)e^{-\\Phi_2(s)/D},\\qquad \\operatorname{Re}\\Phi_1(s)=\\operatorname{Re}\\Phi_2(s),\\qquad Z_c(T;s)=0.$$",
  "$$\\widehat Z_c(T;s)=\\frac{1}{K}\\sum_{r=1}^{K}\\exp\\left[-\\frac{\\mathcal{A}_r}{D}-s\\mathcal{R}_r\\right],\\qquad g(T;s)=-\\frac{1}{K}\\log|\\widehat Z_c(T;s)|^2.$$"
 ],
 "id": 2998,
 "implementation": "Integrate this as a diagnostic wrapper around SGD, Adam, or Langevin training, not initially as a replacement optimizer. At a checkpoint, clone the current parameters into $K=32$ to $128$ replicas. Evolve each replica for $M$ additional steps with independent minibatch ordering and optional Gaussian perturbations, using $\\theta_{k+1}=\\theta_k-\\eta g_k+\\sqrt{2D\\eta}\\xi_k$. Define the initial basin $B_0$ by a parameter-distance threshold and a loss window. For replicas returning to $B_0$, record the trajectory action $\\mathcal A_r$ and return observable $\\mathcal R_r$, such as final squared distance from the basin center or final loss increase. Evaluate $\\widehat Z_c(T;s)$ on a grid of complex probes $s=u+iv$ using log-sum-exp scaling and complex arithmetic. Estimate the smallest $|v|$, denoted $v_*$, at which the complex sum approaches zero, and monitor the return rate $g(T;u)$. The paper supplies the path-competition mechanism; noise scale, basin definition, action approximation, and saddle costs are estimated empirically. First test a two-dimensional double-well loss, then a small MNIST MLP, against unmodified SGD with equal compute. The prediction is that near a basin transition, $g(T;u)$ develops a sharp peak or kink, $v_*$ decreases as $D$ decreases, and two trajectory clusters have approximately equal action at the transition: $|\\mathcal A_1-\\mathcal A_2|/\\max(\\mathcal A_1,\\mathcal A_2)\u003c0.2$. Trigger a learning-rate reduction when $v_*\u003c0.1$ or when the return-rate slope exceeds a baseline threshold. If no cluster exchange, kink, or predictive relation between $v_*$ and basin occupancy occurs, reject the mechanism.",
 "math_summary": "The quantum generating function is the return amplitude $Z_q(t)=\\langle\\psi_0|e^{-iHt/\\hbar}|\\psi_0\\rangle$, where $H$ is the Hamiltonian, $t$ is time, and $\\psi_0$ is the initial state. Its classical weak-noise analogue is a path integral $Z_c(T;s)=\\int\\mathcal{D}x\\exp[-\\mathcal{A}[x]/D-s\\mathcal{R}[x]]$, where $T$ is the horizon, $D$ is noise strength, $\\mathcal{A}[x]$ is the stochastic action, $s$ is a real or complex probe parameter, and $\\mathcal{R}[x]$ is a return observable. In the weak-noise limit, competing trajectory families give $Z_c\\approx C_1(s)e^{-\\Phi_1(s)/D}+C_2(s)e^{-\\Phi_2(s)/D}$, with transition boundary $\\operatorname{Re}\\Phi_1=\\operatorname{Re}\\Phi_2$ and Fisher zeros satisfying $Z_c=0$. For Langevin-like training, $\\theta_{k+1}=\\theta_k-\\eta\\nabla L(\\theta_k)+\\sqrt{2D\\eta}\\,\\xi_k$, with learning rate $\\eta$, loss $L$, and Gaussian noise $\\xi_k$. A discrete action estimator is $\\mathcal{A}_r=\\sum_k\\|\\theta_{k+1}^{(r)}-\\theta_k^{(r)}+\\eta\\nabla L(\\theta_k^{(r)})\\|^2/(4D\\eta)$.",
 "math_tags": [
  "dynamical-systems",
  "statistical-mechanics",
  "stochastic-processes",
  "bifurcations"
 ],
 "ml_areas": [
  "optimizer",
  "training-dynamics",
  "scheduler"
 ],
 "paper": {
  "arxiv_id": "2609.01197",
  "arxiv_url": "https://arxiv.org/abs/2609.01197",
  "summary_what_math_gives_to_ml": "The paper provides a transferable mechanism for detecting dynamical phase transitions through competition between return trajectories, rather than equilibrium free energies. Its central objects are a dynamical generating function, Fisher zeros that approach the physical time axis, and an order parameter distinguishing competing trajectory families. For neural networks, the most promising transfer is to treat an ensemble of stochastic optimization or inference trajectories as a finite-noise path ensemble and monitor a complexified return generating function. A zero approach or saddle-weight crossing could signal an abrupt change in basin occupancy and trigger learning-rate or noise scheduling.",
  "title": "Dynamical phase transitions for single particles in the semiclassical and weak noise limits",
  "year": "2026"
 },
 "ratings": {
  "difficulty": 6,
  "novelty": 8,
  "usefulness": 6
 },
 "solves": [
  "stability",
  "generalization",
  "sample-efficiency"
 ],
 "title": "Fisher-Zero Monitor for Stochastic Training",
 "url": "https://synthcore.org/idea/2998/fisher-zero-monitor-for-stochastic-training",
 "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
   }
  }
 }
}