Paper Radar Archive
The archive preserves the discovery metadata that should outlive the rolling Daily Paper Radar. Records are keyed by the stable base arXiv identifier, so a revision updates one paper history instead of creating a duplicate paper.
Loading the archive manifest…
No archive records match these filters.
Need the structured data?Open the JSON archive manifest. Each month in that manifest links to a stable, static JSON partition.
Reading the record
Section titled “Reading the record”Each paper keeps its first and most recent discovery times, observed arXiv versions,
discovery lanes, and suggested catalog routes. Month partitions are based on when a
paper was first discovered. Later revisions remain attached to that original
record, while last seen and the version history show subsequent activity.
Historical records keep a compact abstract snippet for search and triage, not a mirrored full abstract or PDF. Follow the arXiv link for the canonical current record.
The route labels are discovery suggestions, not verified scientific classifications. They help a maintainer find the right reading chapter; they do not automatically change the curated catalog.
Triage states
Section titled “Triage states”-
Unreviewed — discovered but not yet assessed by a maintainer.
-
Shortlisted — worth a closer methodological and evidence review.
-
Promoted to catalog — verified and assigned a canonical catalog identity. The archive links directly to that exact catalog record.
-
Dismissed — reviewed but not selected for the maintained reading routes. The record stays in the archive so the decision is not silently forgotten.
Promotion is intentionally human-confirmed. A promoted paper remains one archive record linked to one canonical catalog record; it is not copied into a separate “monthly update” collection.
Retention and machine-readable access
Section titled “Retention and machine-readable access”The archive is split into monthly static JSON files. The page fetches only the newest month by default, and loads other months—or every month—only when selected. This keeps the site responsive as the history grows.
The JSON archive manifest lists every available partition, its record count, the query version, source status, and generation timestamps. It is part of the published site and can be used for audits or downstream research tooling.