Takes a field's published claims, reruns them on one common dataset under one protocol, and reports which ones reproduce and which competing explanations the data can actually tell apart.
AI on Alzheimer Disease
Frozen field-scale reconstruction on common empirical ground.
—
Study reconstruction
Validated field-scale applications
AI on Alzheimer Disease95,756 publications · 10,884 data-bearing rows · 183 recurring configurations · 204 executable signatures · ≈1.224 million scheduled model operations.
Common-substrate reconstruction327 matched papers: median reported AUC 0.860 vs matched common-substrate AUC 0.678. Best grouped six-marker panel AUC 0.845.
AI on AI527,520 eligible post-transformer papers screened · 300,000 bounded analysis corpus · 219,052 empirically classifiable papers.