M5 · Uncertainty and validation
Bounded research deliverable accepted; full uncertainty campaign incomplete
all 33 original cases reach five seconds at all three maximum-step settings: 99 of 99 valid trajectories and 264 of 264 passing endpoint comparisonsC-305. the largest changes are 0.219% for coarse-to-middle and 0.0423% for middle-to-fine, both below the 1% acceptance limitC-306. The result uses the new continuous closure and strict research acceptance protocol (the new research baseline uses a continuous guarded closure, Newton trials accepted only after residual decrease, and adaptive rollback of the complete state and auxiliary dataC-304).
Result narrative · Technical protocol. The owner accepts the frozen closure and grants physicist sign-off for the reduced bounded-research scope (OQ-52).
Current uncertainty implementation
The source-bound driver uses the adopted continuous baseline and OQ-51 calibration. three selected five-second trajectories pass the raw audit; only one of ten finite-difference controls survives the derivative timeoutC-327. the 18,432-trajectory design projects roughly 1,710–1,877 node-hours in the measured worker layout, before derivatives and recoveryC-328. Full sampling, sensitivity and derivative agreement remain unestablished. OQ-52 closes and accepts the implementation, evidence and bounded pilot as the reduced research deliverable. Full UQ is excluded from that completion scope and requires a separate future decision. Technical scope.
Scope
OQ-47 selects one ITER hybrid scenario, independent declared input priors, the transport member and a coherent transport-discrepancy factor. The three surrogate modules stay behind their Tier A fallbacks. This is sensitivity analysis under stated priors, not experimentally validated prediction uncertainty.
Historical uncertainty result and gate
The historical Monte Carlo and sensitivity trajectories complete (all 256 retained Monte Carlo outputs are finite with zero native linear-step error codes, and 2048 sensitivity evaluations drop no rows; those codes do not establish nonlinear-residual or temporal convergenceC-292). Their zero native linear-step error codes do not establish nonlinear-residual or temporal convergence. Every linearised standard-deviation check is flagged nonlinear (all four outputs fail the fixed 10% standard-deviation cross-check; disagreements are 46.2%, 43.4%, 43.1% and 59.7% for fusion power, thermal energy, central electron temperature and central ion temperatureC-287), so Monte Carlo carries the historical uncertainty report. Hybrid conformal coverage passes the fixed pooled rule in both directions (hybrid calibration covers 0.914/0.915/0.904 of the older hybrid test states; the reverse audit covers 0.887/0.869/0.900; all exceed the fixed 0.85 pooled barC-285); per-run coverage still varies. A separate conditional Bayesian reference-label calculation agrees with its exact posterior (four Metropolis chains give transport-error factor mean -0.01978, standard deviation 0.01992 and split R-hat 1.00266; exact conditional moments are -0.02033 and 0.01984C-288).
The earlier hard-switch baseline failed both independently audited finer ladders under the unchanged temporal criterion (maximum paired changes remain 19.2654% and 1.9773% across the two refinements, exceeding the fixed 1% limit; all 99 retained rows are finite with zero native linear-step error codes and no exclusionsC-295, the additional finer ladder has maximum paired changes of 1.7092% and 1.8560%, above the unchanged 1% criterion; all 99 assessed rows remain valid and no supported replacement step is establishedC-296). The next full ladder also fails after its first pair passes (the first pair stays below 1%, but the finer pair changes fusion power by up to 4.1372% and fails three quantities; all 99 assessed rows remain valid, so no replacement step is supportedC-298), retaining the same cases, priors, members, physics and endpoint. No supported replacement step or final refined uncertainty study is established. Final propagation also requires a complete design that fits the remaining compute authority.
The report attributes the available solver, transport and EC evidence, and states the unquantified pedestal and equilibrium discrepancy (K-035). The physical discrepancy terms cannot be added into a validated fusion-power error bar. Physicist sign-off remains the gate.
The next full independent ladder audit passes source, design, ancestry, raw/trace and historical-overlap checks, but temporal acceptance still fails (the first pair stays below 1%, but the finer pair changes fusion power by up to 4.1372% and fails three quantities; all 99 assessed rows remain valid, so no replacement step is supportedC-298). The first pair passes; only case20/member19 exceeds 1% in the second, failing fusion power, thermal energy and electron temperature. Its EC and pedestal decisions match at shared times while transport decisions differ. No causal attribution, supported step or final refined uncertainty result follows. The conditional parallel UQ prototype remains unexecuted.
The bounded investigation closes scientifically unresolved after independent stopping review. Further refinement and a sixteen-worker scheduling path remain technically credible, but their complete cost is unmeasured. The stop preserves unused authority and every failed case; it does not claim exhausted compute or impossible convergence.
Earlier numerical diagnostics
The three-case domain/balance observer identifies sustained fallback occupancy and late coefficient-lag residuals in the failing case ([the failing case rejects face seven on 24.766% of coarse and 82.539% of fine accepted states over (3,5] seconds; its median normalized global electron balance residual rises from 0.031813% to 0.962581%]{.claim}C-299). A separately frozen internal observer preserves the original trajectory and establishes different last-guess and accepted-state guard choices ([the failing case has different last-guess and accepted-state branches on 39.453% of coarse and 73.145% of fine steps over (3,5] seconds; during (3.5,4.5] seconds all fine accepted states fall back while 84.961% of last guesses select the surrogate]{.claim}C-300). Its fixed-state branch term is substantial locally but cancels in the signed global residual (the fine failing case has branch-term norm ratios of 0.723 for electron temperature and 0.982 for ion temperature; the signed global branch term cancels and the remaining coefficient lag carries the net defectC-301). These diagnostics do not establish a persistent iteration cycle, attribute the endpoint power difference or support a remedy. See the technical explanation.
At that historical stage, the extra-iterate checkpoint protocol was prepared but unimplemented and unsubmitted. The later continuous baseline now passes the complete comparison; full uncertainty/sensitivity execution remains incomplete; OQ-52 subsequently grants physicist sign-off for the reduced scope.
Current numerical baseline
The complete original case design now passes its declared five-second numerical protocol (all 33 original cases reach five seconds at all three maximum-step settings: 99 of 99 valid trajectories and 264 of 264 passing endpoint comparisonsC-305). Both adjacent adaptive-cap comparisons pass for every case and every quantity (the largest changes are 0.219% for coarse-to-middle and 0.0423% for middle-to-fine, both below the 1% acceptance limitC-306), and all accepted steps pass the physical/native residual gates (maximum accepted physical-channel L1 is 1.76 × 10−9 against a 10−8 gate, and native mean L1 is 1.01 × 10−14 against 10−10; no crash occursC-307). Rejected proposals and rollback checks are retained (372,141 accepted proposals and 2,659 rejected proposals are retained in 23,100 raw chunks; rejected-step rollback and retry schedules pass reconstructionC-308).
This is a changed continuous transport closure, not a retrospective pass for the historical hard-switch equations. The original case design is preserved while the method is explicitly versioned. A subsequent uncertainty/sensitivity study must use this baseline deliberately; the prior report has not been silently recomputed. Smoothing-width robustness and physicist acceptance remain open.