Answers are grounded in this paper and cite their sources.
Data descriptor for the SC-25 dataset (preprint). Abstract: identical to 04-manuscript/index.qmd at release. Competing interests: G. Gonçalves is Vice President of Industrial AI Lab (non-profit) and owns SmartHow, a company that provides AI solutions for industry. This work uses AI-assisted methods, an area in which SmartHow operates (relation to this work: yes, AI). No other competing interests are declared. Note on the CertBasis-Bench hold-out: its seal protects only against exposure through that benchmark's own evaluation pipeline. The Federal Register text is public and SC-25 makes the answers of most sealed items derivable, so a future confirmatory test needs a prospective hold-out (special conditions issued after this release and after each model's training cutoff) or a fresh seal; the SHA-256 commitment still proves which items were sealed, and when.
No takes yet. Share an insight, caveat, or question.
Gabriel Henrique Antunes Goncalves (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: