{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:3NOOSX3OXFEFQ4K2Q7ZTSVX66D","short_pith_number":"pith:3NOOSX3O","schema_version":"1.0","canonical_sha256":"db5ce95f6eb94858715a87f33956fef0f2d8ee74b860709915cb193bfa020c5e","source":{"kind":"arxiv","id":"2503.17405","version":2},"attestation_state":"computed","paper":{"title":"Efficiently Vectorized MCMC on Modern Accelerators","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG","stat.CO","stat.ML"],"primary_cat":"cs.MS","authors_text":"Hugh Dance, Peter Orbanz, Pierre Glaser, Ryan Adams","submitted_at":"2025-03-20T16:07:14Z","abstract_excerpt":"With the advent of automatic vectorization tools (e.g., JAX's $\\texttt{vmap}$), writing multi-chain MCMC algorithms is often now as simple as invoking those tools on single-chain code. Whilst convenient, for various MCMC algorithms this results in a synchronization problem -- loosely speaking, at each iteration all chains running in parallel must wait until the last chain has finished drawing its sample. In this work, we show how to design single-chain MCMC algorithms in a way that avoids synchronization overheads when vectorizing with tools like $\\texttt{vmap}$ by using the framework of finit"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2503.17405","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.MS","submitted_at":"2025-03-20T16:07:14Z","cross_cats_sorted":["cs.LG","stat.CO","stat.ML"],"title_canon_sha256":"dbb12b60043041976e67957b55564066358c20f84cb8659a201ca11dbf6ffbd4","abstract_canon_sha256":"5f1f31c42bd71e667dd72ee6e148519462355e3b84b70a15ca045c238200848f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:30:42.594482Z","signature_b64":"hQUZ96Egy0DqqslBtwer1CClmq97DOgPVhHPuENRjPfa0eLJ6LBkbMnsTem+VNSRV1WRXan8dv3cs4W4SiUQDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"db5ce95f6eb94858715a87f33956fef0f2d8ee74b860709915cb193bfa020c5e","last_reissued_at":"2026-07-05T11:30:42.593949Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:30:42.593949Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Efficiently Vectorized MCMC on Modern Accelerators","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG","stat.CO","stat.ML"],"primary_cat":"cs.MS","authors_text":"Hugh Dance, Peter Orbanz, Pierre Glaser, Ryan Adams","submitted_at":"2025-03-20T16:07:14Z","abstract_excerpt":"With the advent of automatic vectorization tools (e.g., JAX's $\\texttt{vmap}$), writing multi-chain MCMC algorithms is often now as simple as invoking those tools on single-chain code. Whilst convenient, for various MCMC algorithms this results in a synchronization problem -- loosely speaking, at each iteration all chains running in parallel must wait until the last chain has finished drawing its sample. In this work, we show how to design single-chain MCMC algorithms in a way that avoids synchronization overheads when vectorizing with tools like $\\texttt{vmap}$ by using the framework of finit"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.17405","kind":"arxiv","version":2},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2503.17405/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2503.17405","created_at":"2026-07-05T11:30:42.594009+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.17405v2","created_at":"2026-07-05T11:30:42.594009+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.17405","created_at":"2026-07-05T11:30:42.594009+00:00"},{"alias_kind":"pith_short_12","alias_value":"3NOOSX3OXFEF","created_at":"2026-07-05T11:30:42.594009+00:00"},{"alias_kind":"pith_short_16","alias_value":"3NOOSX3OXFEFQ4K2","created_at":"2026-07-05T11:30:42.594009+00:00"},{"alias_kind":"pith_short_8","alias_value":"3NOOSX3O","created_at":"2026-07-05T11:30:42.594009+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3NOOSX3OXFEFQ4K2Q7ZTSVX66D","json":"https://pith.science/pith/3NOOSX3OXFEFQ4K2Q7ZTSVX66D.json","graph_json":"https://pith.science/api/pith-number/3NOOSX3OXFEFQ4K2Q7ZTSVX66D/graph.json","events_json":"https://pith.science/api/pith-number/3NOOSX3OXFEFQ4K2Q7ZTSVX66D/events.json","paper":"https://pith.science/paper/3NOOSX3O"},"agent_actions":{"view_html":"https://pith.science/pith/3NOOSX3OXFEFQ4K2Q7ZTSVX66D","download_json":"https://pith.science/pith/3NOOSX3OXFEFQ4K2Q7ZTSVX66D.json","view_paper":"https://pith.science/paper/3NOOSX3O","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.17405&json=true","fetch_graph":"https://pith.science/api/pith-number/3NOOSX3OXFEFQ4K2Q7ZTSVX66D/graph.json","fetch_events":"https://pith.science/api/pith-number/3NOOSX3OXFEFQ4K2Q7ZTSVX66D/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3NOOSX3OXFEFQ4K2Q7ZTSVX66D/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3NOOSX3OXFEFQ4K2Q7ZTSVX66D/action/storage_attestation","attest_author":"https://pith.science/pith/3NOOSX3OXFEFQ4K2Q7ZTSVX66D/action/author_attestation","sign_citation":"https://pith.science/pith/3NOOSX3OXFEFQ4K2Q7ZTSVX66D/action/citation_signature","submit_replication":"https://pith.science/pith/3NOOSX3OXFEFQ4K2Q7ZTSVX66D/action/replication_record"}},"created_at":"2026-07-05T11:30:42.594009+00:00","updated_at":"2026-07-05T11:30:42.594009+00:00"}