{
 "artifacts": null,
 "category": "dynamics",
 "description": "Add an explicit transport-delay state to a recurrent neural network, state-space model, or learned optimizer whenever actions, gradients, or control inputs arrive after a fixed delay. Use the queued inputs to construct a finite-horizon predictor state and apply the neural transition or controller to that predicted state rather than to the stale state. The design transfers the paper's delay-as-transport-PDE and backstepping-to-stable-target strategy into a differentiable predictor with an observable contraction margin.",
 "formulas_latex": [
  "$$q_t(s,t)+q_s(s,t)=0,\\qquad q(0,t)=u(t),\\qquad q(\\tau,t)=u(t-\\tau).$$",
  "$$\\hat{x}_{k+d}=A^d x_k+\\sum_{j=0}^{d-1}A^{d-1-j}B u_{k+j-d},\\qquad u_k=K\\hat{x}_{k+d}.$$",
  "$$\\rho(A+BK)\u003c1\\quad\\text{or}\\quad (A+BK)^\\top P(A+BK)-P\\preceq-\\alpha P.$$",
  "$$\\|x_k\\|\\leq C\\lambda^k\\|x_0\\|+\\gamma\\sup_{j\u003ck}\\|e_j\\|,\\qquad \\gamma L_e\u003c1.$$"
 ],
 "id": 3097,
 "implementation": "Integrate this at the hidden-state update of a GRU-free RNN, a diagonal-plus-low-rank SSM, or a learned optimizer whose parameter update is delayed by asynchronous workers. Maintain a FIFO queue containing the last d inputs or gradients. At every step, estimate the local Jacobians A and B of the neural transition f(x,u) using automatic-differentiation Jacobian-vector products; for the first MVP, use a frozen linearization estimated once per minibatch. Compute the predictor with repeated matrix-vector products: set p=x, then apply p \u003c- A@p+B@queued_input[j] for each future slot in chronological order. Feed p, rather than the stale x, into the controller u=pi_theta(p), or use a linear K during the stability experiment. Pseudocode is: enqueue the newest input; p=x; for j=0,...,d-1, p=A@p+B@queued_input[j]; u=controller(p); execute u after the prescribed delay; update the network normally. The transport-PDE part supplies the queue interpretation; A, B, and the residual prediction error are estimated empirically. First test on a 2-layer tanh RNN learning the delayed-copy task and on an SSM trained on sequential MNIST, comparing ordinary delayed feedback with the predictor for d in {1,4,16,64}. Estimate rho(A+BK) by power iteration along trajectories and sweep the controller gain until it crosses one. The predicted signature is a sharp divergence boundary near rho(A+BK)=1, largely independent of d for the compensated model, while the uncompensated model's stable delay range shrinks as d increases. Also inject bounded Gaussian transition noise: state norms should remain bounded below the measured small-gain threshold and grow rapidly when the estimated product gamma times L_e exceeds one.",
 "math_summary": "Represent a delay of length \\(\\tau\\) by the transport equation \\(q_t(s,t)+q_s(s,t)=0\\), for \\(s\\in[0,\\tau]\\), with boundary condition \\(q(0,t)=u(t)\\); therefore \\(q(\\tau,t)=u(t-\\tau)\\). For a locally linearized neural state \\(x_{k+1}=Ax_k+Bu_{k-d}+w_k\\), the delay-compensated predictor is \\(\\hat{x}_{k+d}=A^d x_k+\\sum_{j=0}^{d-1}A^{d-1-j}B u_{k+j-d}\\), where \\(d\\) is the integer delay, \\(A=\\partial f/\\partial x\\), \\(B=\\partial f/\\partial u\\), and \\(w_k\\) is model error. Apply \\(u_k=K\\hat{x}_{k+d}\\). The target closed-loop matrix is \\(A+BK\\), and exponential stability requires \\(\\rho(A+BK)\u003c1\\), or, more robustly, \\((A+BK)^\\top P(A+BK)-P\\preceq-\\alpha P\\) for some \\(P\\succ0\\) and \\(\\alpha\u003e0\\). With prediction error \\(e_k\\), an ISS bound has the form \\(\\|x_k\\|\\leq C\\lambda^k\\|x_0\\|+\\gamma\\sup_{j\u003ck}\\|e_j\\|\\); the small-gain requirement is \\(\\gamma L_e\u003c1\\), where \\(L_e\\) bounds the error feedback gain.",
 "math_tags": [
  "control-theory",
  "dynamical-systems",
  "pde",
  "linear-algebra",
  "optimization"
 ],
 "ml_areas": [
  "rnn",
  "ssm",
  "optimizer",
  "training-dynamics"
 ],
 "paper": {
  "arxiv_id": "2609.03345",
  "arxiv_url": "https://arxiv.org/abs/2609.03345",
  "summary_what_math_gives_to_ml": "The paper offers a constructive delay-compensation mechanism for hyperbolic systems: represent an arbitrarily large input delay as a transport PDE, then use a backstepping transformation to map the delayed plant into an exponentially stable target system. Its transferable asset is not the traffic model itself, but the combination of a predictor-like Volterra transformation, a stable target dynamics, and an ISS small-gain condition for residual disturbances. A neural analogue is a delay-compensated recurrent or state-space controller that predicts the hidden state under a queued sequence of stale actions, while explicitly monitoring a contraction or small-gain margin. The strongest falsifiable prediction is that compensation should preserve the stability boundary as delay grows, whereas an uncompensated recurrent controller becomes unstable once the effective delayed-loop gain exceeds one.",
  "title": "Traffic Congestion Control for ARZ Model with an Arbitrarily Large Input Delay",
  "year": "2026"
 },
 "ratings": {
  "difficulty": 6,
  "novelty": 6,
  "usefulness": 7
 },
 "solves": [
  "stability",
  "accuracy"
 ],
 "title": "Transport-PDE Predictor for Delayed Neural State Updates",
 "url": "https://synthcore.org/idea/3097/transport-pde-predictor-for-delayed-neural-state-updates",
 "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
   }
  }
 }
}