Orca OSDI '22 paper - NO arXiv preprint exists; the arXiv API returns zero results for this title

HONEST NEGATIVE recorded so the gap is not re-litigated: the OSDI'22 Orca serving paper has NO arXiv version. Verified 2026-10-04 by querying export.arxiv.org/api/query with search_query ti:"Orca: A Distributed Serving System for Transformer-Based Generative Models" (0 results), ti:"Orca" AND all:"seqlib" (0), ti:"Distributed Serving System for Transformer-Based" (0), all:"Orca" AND all:"iteration-level scheduling" (0). The only arXiv paper titled Orca is 2306.02707 (Microsoft's distillation paper, a DIFFERENT paper). Consequence: the arXiv ID field for the serving paper must be null, and the repo's existing 'orca' paper record (2306.02707) is a different work that happens to share the name. Affiliations and pagination from the USENIX page citation_* meta tags: Gyeong-In Yu, Joo Seong Jeong (Seoul National University); Geon-Woo Kim (FriendliAI and Seoul National University); Soojeong Kim (FriendliAI); Byung-Gon Chun (FriendliAI and Seoul National University). OSDI 22 pp. 521-538, ISBN 978-1-939133-28-1.