Plan the fewest, most-informative pulses to validate — accelerate commissioning.
The AI-Native S.M.A.R.T. Generator Master Blueprint — eight layers (L0→L7), one control stack, wired to both machines. Telemetry rises in microseconds; control descends the same path.
Category: C · mathematics · Plugs into: L4 / L5 · Horizon: FOAK · Status: on the roadmap — not yet built
What it is
Rather than a fixed test matrix, choose each next pulse to maximize the information gained about the models — active learning for a fusion machine.
The method
Bayesian optimization with an expected-information-gain acquisition over the pulse-parameter space; the twin proposes the next most-informative shot.
Why it matters
Shots are expensive and commissioning is the schedule's long pole; this minimizes shots-to-validate. Plugs into L4/L5 — the workflow engine and the operations copilot.