ESC

Source Ingest Queue

This is the master queue for source collection and ingest.

Acquisition lanes

  1. Initial Corpus Batch 01
  2. Ilya Top 30
  3. Autonomous Driving Seminal Papers
  4. Vla And Driving
  5. Llm Seminal Papers

Intake workflow

For each source:

  1. Save the raw artifact to raw/inbox/ or raw/papers/.
  2. Normalize filename to year-short-title.pdf when possible.
  3. Create or update the matching wiki/sources/ page.
  4. Route updates into concept pages and comparisons.
  5. Record unresolved questions that the source creates.

Notes

  • Citation-count thresholds should be verified at ingest time.
  • For source discovery, use paper graphs and citations to expand outward from canonical papers rather than scraping arbitrary lists.
  • If AlphaXiv-specific tooling is unavailable, use arXiv, OpenReview, Semantic Scholar, OpenAlex, and project pages.
  • The initial real paper set now lives in Initial Corpus Batch 01.