INTELLIGENCE Live
Research Library
The document and knowledge corpus: research papers, books, library items, and the embedding layer that makes it all semantically searchable. Includes the arXiv mirror.
- Sources
- arXiv·research feeds·book hoard·embedded corpus
- Volume
- 2.7M+ embeddings, 290K+ library items
- Depth
- Full documents, books, papers, vector embeddings
- Freshness
- pending refresh
Backfill
continuous stream, no fixed target
Tables and row counts
| Table | Rows | Notes |
|---|---|---|
| embeddings | 2,762,484 | vector embeddings |
| library_items | pending | |
| research_documents | 261,949 | |
| library_books | 28,444 | ebook subset |
| arxiv (Meili) | 1,030,000 | approximate, MeiliSearch index |
Exact counts are verified against the live cluster. Fields shown as pending will be filled by the producer on the next quiet-box refresh.
What a record looks like
embeddings backs semantic retrieval across the corpus. arXiv (~1.03M papers) is searchable via MeiliSearch. library_books is the 24K+ ebook subset.