Randomized trial shows enhanced clinical discrimination and drift detection in edge deployments, indicating improved model readiness.
A 100-study pre-registered Modal cloud A100 stress test of the spiral-domain H-pipeline (studies 1741–1840, fifteen sub-campaigns α through ο, 70.0 percent SUPPORTED). Three Outcome A reframing findings: (θ) 5-class clinical macro-AUC 0.657 beats Chronos 0.642 at 511× lower bandwidth; (μ) drift / out-of-source detection AUC = 1.000 at 30-microsecond latency; (ξ) host-CPU streaming p99 = 0.385 ms matches Modal A100 p99 = 0.389 ms, refuting the "GPU required" objection to edge deployment. Foundation-model field rank extends Paper 16's Chronos-only baseline: H-pipeline = 1 BPE, Moirai = 511.46, MOMENT = 511.94, TimesFM = 512, Chronos = 900, naive cl100k_base = 1024. Six honest Outcome C disclosures preserve pre-registration discipline. Companion deposits in this Zenodo Community (spiral-domain-encoder-campaign): · Paper 8 — 10.5281/zenodo.20466035 · Paper 9 — 10.5281/zenodo.20470763 · Paper 10 — 10.5281/zenodo.20480183 · Paper 11 — 10.5281/zenodo.20495976 · Paper 12 — 10.5281/zenodo.20499453 · Paper 13 — 10.5281/zenodo.20595834 · Paper 14 v3 — 10.5281/zenodo.20596597 · Paper 15 — 10.5281/zenodo.20596931 · Paper 16 — 10.5281/zenodo.20602959 · Paper 17 — [10.5281/zenodo.20634763](https://doi.org/10.5281/zenodo.20634763) ← this deposit Related U.S. Provisional Patent Applications (all previously filed; no new provisional required for this paper): Parent H (64/080,096), Parent I (64/084,807), Parent J (64/084,817), Parent K (64/084,821), Parent L (64/084,827). 12-month non-provisional conversion deadline: 2027-06-11. Licensing inquiries: Randolph James Ferlic, M.D., randolphf@fieldstoneanalyticsllc.com.
No takes yet. Share an insight, caveat, or question.
Ferlic et al. (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: