[ayush1298]

refs #5703. 1st part to update citations before #5722
Updates 202 citations that cited the arXiv preprint of a paper that has since been published.

• BibTeX comes from the ACL Anthology, doi.org, OpenReview or a Crossref title match. Ten entries were filled in by hand from the proceedings pages: CLIP, Whisper and WebLINX (PMLR), two ICLR papers, three TREC overviews, AV-SpeakerBench and DaLAJ. Every citation keeps its key.
• Entries without a DOI keep their arXiv eprint, so the citation check can verify them by id instead of by title search.
• li2023making (Llama2Vec): Chaofan Li is listed first, as in Crossref and arXiv; the ACL Anthology lists Zheng Liu first.
• citation_allowlist.json: adds BEIR-NL (the published version lists the authors in a different order from arXiv) and Flan-T5 (JMLR 2024, arXiv 2022). The pre-commit hook re-indented the file to 4 spaces.
• citation_cache.json: updated by make citation-check, which passes.
• tests/test_get_tasks.py: the duplicate-citation test ignores BibTeX case braces (e.g. {BEIR}-{NL}) when comparing titles, so the ACL Anthology entries are not reported as duplicates.
• reference links that pointed to the arXiv version of an updated paper now point to the published version (ACL Anthology, PMLR, OpenReview, NeurIPS proceedings, or the DOI): 226 links in 109 files.

Not changed:

• MS MARCO still cites the 2018 arXiv revision (Bajaj et al.). The 2016 CoCo@NIPS paper is an earlier version with different authors.
• Jina Embeddings 2 stays on arXiv; as far as it was never published.

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

An organization owner can choose the repository organization as the billing source for reviews requested by members with a paid Copilot license. Learn about Copilot code review billing.

[KennethEnevoldsen]

spot check of about 10 papers all I saw seems correct

[ayush1298]

Should I merge this one? @Samoed