{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:QZ4WHRMQDWXX5FA5Z2JYLXE6OJ","short_pith_number":"pith:QZ4WHRMQ","schema_version":"1.0","canonical_sha256":"867963c5901daf7e941dce9385dc9e727de201bc32afe52fc95521a8748eb018","source":{"kind":"arxiv","id":"2205.09707","version":1},"attestation_state":"computed","paper":{"title":"PLAID: An Efficient Engine for Late Interaction Retrieval","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.IR","authors_text":"Christopher Potts, Keshav Santhanam, Matei Zaharia, Omar Khattab","submitted_at":"2022-05-19T17:19:31Z","abstract_excerpt":"Pre-trained language models are increasingly important components across multiple information retrieval (IR) paradigms. Late interaction, introduced with the ColBERT model and recently refined in ColBERTv2, is a popular paradigm that holds state-of-the-art status across many benchmarks. To dramatically speed up the search latency of late interaction, we introduce the Performance-optimized Late Interaction Driver (PLAID). Without impacting quality, PLAID swiftly eliminates low-scoring passages using a novel centroid interaction mechanism that treats every passage as a lightweight bag of centroi"},"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":"2205.09707","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2022-05-19T17:19:31Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"dad1a2bfcde86c5ec00e33478b1c41edafe8ed33d89b41ede274aa9fa2dca223","abstract_canon_sha256":"ad1bf4cb677a058b1e8c5e4cb2ebc5b2a63b315314a49e81fda3aabcca61aa4e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:24:47.759748Z","signature_b64":"PZ2YjFoQvqPolWXGX6c9FU5OPS70kh2W6A5qeiq8EKWWgj0YrL2qUlVXEwFNp80zOEn5kCqqhZIoBEVnmKnzAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"867963c5901daf7e941dce9385dc9e727de201bc32afe52fc95521a8748eb018","last_reissued_at":"2026-07-05T04:24:47.759256Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:24:47.759256Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"PLAID: An Efficient Engine for Late Interaction Retrieval","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.IR","authors_text":"Christopher Potts, Keshav Santhanam, Matei Zaharia, Omar Khattab","submitted_at":"2022-05-19T17:19:31Z","abstract_excerpt":"Pre-trained language models are increasingly important components across multiple information retrieval (IR) paradigms. Late interaction, introduced with the ColBERT model and recently refined in ColBERTv2, is a popular paradigm that holds state-of-the-art status across many benchmarks. To dramatically speed up the search latency of late interaction, we introduce the Performance-optimized Late Interaction Driver (PLAID). Without impacting quality, PLAID swiftly eliminates low-scoring passages using a novel centroid interaction mechanism that treats every passage as a lightweight bag of centroi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2205.09707","kind":"arxiv","version":1},"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/2205.09707/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":"2205.09707","created_at":"2026-07-05T04:24:47.759310+00:00"},{"alias_kind":"arxiv_version","alias_value":"2205.09707v1","created_at":"2026-07-05T04:24:47.759310+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2205.09707","created_at":"2026-07-05T04:24:47.759310+00:00"},{"alias_kind":"pith_short_12","alias_value":"QZ4WHRMQDWXX","created_at":"2026-07-05T04:24:47.759310+00:00"},{"alias_kind":"pith_short_16","alias_value":"QZ4WHRMQDWXX5FA5","created_at":"2026-07-05T04:24:47.759310+00:00"},{"alias_kind":"pith_short_8","alias_value":"QZ4WHRMQ","created_at":"2026-07-05T04:24:47.759310+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.08673","citing_title":"ClinicalAligner26AM: A Cross-Lingual Aligner for Dataset Translation; Evidences from the MultiClinCorpus Shared Task","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2606.28365","citing_title":"CAMI: Cost-Aware Agent-Guided Multi-Indexing for Semantic Retrieval","ref_index":38,"is_internal_anchor":false},{"citing_arxiv_id":"2604.19566","citing_title":"Diagnosable ColBERT: Debugging Late-Interaction Retrieval Models Using a Learned Latent Space as Reference","ref_index":14,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QZ4WHRMQDWXX5FA5Z2JYLXE6OJ","json":"https://pith.science/pith/QZ4WHRMQDWXX5FA5Z2JYLXE6OJ.json","graph_json":"https://pith.science/api/pith-number/QZ4WHRMQDWXX5FA5Z2JYLXE6OJ/graph.json","events_json":"https://pith.science/api/pith-number/QZ4WHRMQDWXX5FA5Z2JYLXE6OJ/events.json","paper":"https://pith.science/paper/QZ4WHRMQ"},"agent_actions":{"view_html":"https://pith.science/pith/QZ4WHRMQDWXX5FA5Z2JYLXE6OJ","download_json":"https://pith.science/pith/QZ4WHRMQDWXX5FA5Z2JYLXE6OJ.json","view_paper":"https://pith.science/paper/QZ4WHRMQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2205.09707&json=true","fetch_graph":"https://pith.science/api/pith-number/QZ4WHRMQDWXX5FA5Z2JYLXE6OJ/graph.json","fetch_events":"https://pith.science/api/pith-number/QZ4WHRMQDWXX5FA5Z2JYLXE6OJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QZ4WHRMQDWXX5FA5Z2JYLXE6OJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QZ4WHRMQDWXX5FA5Z2JYLXE6OJ/action/storage_attestation","attest_author":"https://pith.science/pith/QZ4WHRMQDWXX5FA5Z2JYLXE6OJ/action/author_attestation","sign_citation":"https://pith.science/pith/QZ4WHRMQDWXX5FA5Z2JYLXE6OJ/action/citation_signature","submit_replication":"https://pith.science/pith/QZ4WHRMQDWXX5FA5Z2JYLXE6OJ/action/replication_record"}},"created_at":"2026-07-05T04:24:47.759310+00:00","updated_at":"2026-07-05T04:24:47.759310+00:00"}