Research & Science
Papers, benchmarks and AI-for-science results, synthesised for builders rather than paper-readers. Benchmarks are first-class entities linked into the model graph.
The entities are in the registry and the pipeline that would cover them is the same one already running elsewhere on this site. What is missing is the golden evaluation set — and the curated creator list the community layer needs. We don't publish a vertical until its classifier is measurably right: a feed that misclassifies is worse than no feed.
The pipeline, schema and design are shared. Turning this on is a config change, not a rebuild. Follow the main stream in the meantime.
ON THE MANIFEST
Tracked in the entity registry, awaiting golden-set sign-off. The registry is append-only: entities are never deleted, only re-statused.
SEE THE REGISTER →