{
 "artifacts": null,
 "category": "architecture",
 "description": "Replace a neural controller's pointwise action outputs over a finite horizon with Bernstein control points whose convex hull satisfies actuator and trajectory constraints. The network predicts the control points, while a robust margin accounts for bounded tracking or model-prediction error, making continuous-time actuator feasibility checkable from finitely many inequalities.",
 "formulas_latex": [
  "$$B_{k,n}(s)=\\binom{n}{k}s^k(1-s)^{n-k},\\qquad u(t)=\\sum_{k=0}^{n}c_kB_{k,n}(t/T),\\qquad B_{k,n}(s)\\geq0,\\quad\\sum_{k=0}^{n}B_{k,n}(s)=1.$$",
  "$$u_{\\min}+\\varepsilon_u\\mathbf{1}\\leq c_k\\leq u_{\\max}-\\varepsilon_u\\mathbf{1}\\quad\\forall k\\ \\Longrightarrow\\ u_{\\min}\\leq u(t)+e_u(t)\\leq u_{\\max}\\quad\\forall t,$$",
  "$$d_k=\\frac{n}{T}(c_{k+1}-c_k),\\qquad \\dot u(t)=\\sum_{k=0}^{n-1}d_kB_{k,n-1}(t/T),\\qquad d_{\\min}\\leq d_k\\leq d_{\\max}.$$",
  "$$\\varepsilon_u=\\|K\\|_\\infty\\varepsilon_e,\\qquad \\|K\\|_\\infty=\\max_i\\sum_j|K_{ij}|.$$"
 ],
 "id": 2919,
 "implementation": "1. Integration point: use an MLP policy head that receives the current observation and predicts $m(n+1)$ control-point values instead of one action. For a receding-horizon controller, execute only the first action or first short segment, then re-predict at the next observation; for offline trajectory generation, decode the complete Bernstein curve. 2. Pseudocode: compute raw control points $C_{raw}=\\mathrm{policy}_\\theta(o)$; estimate or provide an error bound $\\varepsilon_e$; set $\\varepsilon_u=\\|K\\|_\\infty\\varepsilon_e$; project $C_{raw}$ onto $u_{min}+\\varepsilon_u\\leq C\\leq u_{max}-\\varepsilon_u$; compute $D=n(C_{1:}-C_{:-1})/T$ and either project onto slew constraints or add a hinge penalty; evaluate $u(t_j)=B(t_j)C$ on the simulator grid. 3. Computed from the mechanism: Bernstein basis values, control-point bounds, derivative control points, and the convex-hull certificate. Estimated empirically: $\\varepsilon_e$ from a held-out rollout quantile or a conservative disturbance bound, and $K$ from the tracking controller. 4. First experiment: train PPO or model-predictive imitation on a 2D point-mass or pendulum with identical network size, horizon, and environment steps. Compare a standard action head, an unconstrained Bernstein head, and the robust Bernstein head; ablate the error margin and slew constraints. 5. Pre-registered signature: with all control-point margins satisfied, the maximum actuator violation over at least 1000 uniformly sampled times per rollout must be zero up to numerical tolerance $10^{-6}$, including between training-grid points. The measured violation rate must stay below 0.1%; otherwise reject the convex-hull transfer. Removing margin $\\varepsilon_u$ should produce violations once observed tracking error exceeds the removed margin, while the robust head should remain feasible. If violations occur despite valid control-point inequalities, reject the implementation or the assumed actuator/error model.",
 "math_summary": "For horizon $t\\in[0,T]$, define the degree-$n$ Bernstein basis $B_{k,n}(s)=\\binom{n}{k}s^k(1-s)^{n-k}$ with normalized time $s=t/T$ and control trajectory $u(t)=\\sum_{k=0}^{n}c_kB_{k,n}(s)$, where $c_k\\in\\mathbb{R}^m$ are neural-network outputs. Since $B_{k,n}(s)\\geq0$ and $\\sum_kB_{k,n}(s)=1$, every $u(t)$ lies in the convex hull of the control points; therefore componentwise bounds $u_{\\min}\\leq c_k\\leq u_{\\max}$ imply $u_{\\min}\\leq u(t)\\leq u_{\\max}$ for all $t$. Derivative control points are $d_k=n(c_{k+1}-c_k)/T$, giving $\\dot u(t)=\\sum_{k=0}^{n-1}d_kB_{k,n-1}(s)$, so slew-rate limits follow from $d_{\\min}\\leq d_k\\leq d_{\\max}$. If execution has additive tracking error $e_u(t)$ satisfying $\\|e_u(t)\\|_\\infty\\leq\\varepsilon_u$, robust actuator feasibility is guaranteed by $u_{\\min}+\\varepsilon_u\\leq c_k\\leq u_{\\max}-\\varepsilon_u$; more generally, for a feedback correction $K e(t)$ with $\\|e(t)\\|_\\infty\\leq\\varepsilon_e$, use margin $\\varepsilon_u=\\|K\\|_\\infty\\varepsilon_e$.",
 "math_tags": [
  "control-theory",
  "optimization",
  "dynamical-systems",
  "approximation-theory"
 ],
 "ml_areas": [
  "rl",
  "world-model",
  "mlp",
  "training"
 ],
 "paper": {
  "arxiv_id": "2607.23040",
  "arxiv_url": "https://arxiv.org/abs/2607.23040",
  "summary_what_math_gives_to_ml": "The paper offers a constructive way to enforce actuator and geometric constraints before execution: represent a time-varying trajectory with Bernstein polynomials and use their convex-hull property to convert continuous-time bounds into finite linear inequalities on control points. Its second transferable mechanism is robust feasibility under tracking error, where worst-case controller error is converted into a margin in the actuator constraints rather than handled by online re-optimization. A practical neural-network transfer is a Bernstein control-point policy head for finite-horizon control or trajectory generation, optionally followed by a projection layer, with actuator limits enforced directly during training and inference. The key falsifiable prediction is that constraint violations should remain zero over the entire interpolated horizon whenever all control-point inequalities, including the tracking-error margin, are satisfied.",
  "title": "Actuator-Aware Spatiotemporal Tube Synthesis for Temporal Reach-Avoid-Stay Tasks",
  "year": "2026"
 },
 "ratings": {
  "difficulty": 5,
  "novelty": 7,
  "usefulness": 6
 },
 "solves": [
  "stability",
  "accuracy",
  "sample-efficiency"
 ],
 "title": "Bernstein-Safe Policy Head",
 "url": "https://synthcore.org/idea/2919/bernstein-safe-policy-head",
 "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
   }
  }
 }
}