{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:V4D267W4SOMKDMREQVRCLBVOWS","short_pith_number":"pith:V4D267W4","schema_version":"1.0","canonical_sha256":"af07af7edc9398a1b22485622586aeb4b9e095d2dad7be76eb90e4b97e036f3e","source":{"kind":"arxiv","id":"2304.01982","version":3},"attestation_state":"computed","paper":{"title":"Rethinking the Role of Token Retrieval in Multi-Vector Retrieval","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.IR"],"primary_cat":"cs.CL","authors_text":"Iftekhar Naim, Jinhyuk Lee, Ming-Wei Chang, Sai Meher Karthik Duddu, Tao Lei, Vincent Y. Zhao, Zhuyun Dai","submitted_at":"2023-04-04T17:37:06Z","abstract_excerpt":"Multi-vector retrieval models such as ColBERT [Khattab and Zaharia, 2020] allow token-level interactions between queries and documents, and hence achieve state of the art on many information retrieval benchmarks. However, their non-linear scoring function cannot be scaled to millions of documents, necessitating a three-stage process for inference: retrieving initial candidates via token retrieval, accessing all token vectors, and scoring the initial candidate documents. The non-linear scoring function is applied over all token vectors of each candidate document, making the inference process co"},"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":"2304.01982","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-04-04T17:37:06Z","cross_cats_sorted":["cs.IR"],"title_canon_sha256":"ccd38c62fae7ec5ec1ac234629db99d942848daf69ca1855174d561682c08f90","abstract_canon_sha256":"ee5877258e64091bea6361e4734e97bb2c1c5331023cdbe588a5aafef8bb4b7a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:05:48.408710Z","signature_b64":"RwEFj6N9ysSPyqQGJpdwrBf9zcOq9VM3GepHhWKHl/OWZt0he+x5DSGzvaV90kI4AP6zHNvTboBtSXQo7fwhBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"af07af7edc9398a1b22485622586aeb4b9e095d2dad7be76eb90e4b97e036f3e","last_reissued_at":"2026-07-05T08:05:48.408169Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:05:48.408169Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Rethinking the Role of Token Retrieval in Multi-Vector Retrieval","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.IR"],"primary_cat":"cs.CL","authors_text":"Iftekhar Naim, Jinhyuk Lee, Ming-Wei Chang, Sai Meher Karthik Duddu, Tao Lei, Vincent Y. Zhao, Zhuyun Dai","submitted_at":"2023-04-04T17:37:06Z","abstract_excerpt":"Multi-vector retrieval models such as ColBERT [Khattab and Zaharia, 2020] allow token-level interactions between queries and documents, and hence achieve state of the art on many information retrieval benchmarks. However, their non-linear scoring function cannot be scaled to millions of documents, necessitating a three-stage process for inference: retrieving initial candidates via token retrieval, accessing all token vectors, and scoring the initial candidate documents. The non-linear scoring function is applied over all token vectors of each candidate document, making the inference process co"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2304.01982","kind":"arxiv","version":3},"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/2304.01982/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":"2304.01982","created_at":"2026-07-05T08:05:48.408229+00:00"},{"alias_kind":"arxiv_version","alias_value":"2304.01982v3","created_at":"2026-07-05T08:05:48.408229+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2304.01982","created_at":"2026-07-05T08:05:48.408229+00:00"},{"alias_kind":"pith_short_12","alias_value":"V4D267W4SOMK","created_at":"2026-07-05T08:05:48.408229+00:00"},{"alias_kind":"pith_short_16","alias_value":"V4D267W4SOMKDMRE","created_at":"2026-07-05T08:05:48.408229+00:00"},{"alias_kind":"pith_short_8","alias_value":"V4D267W4","created_at":"2026-07-05T08:05:48.408229+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/V4D267W4SOMKDMREQVRCLBVOWS","json":"https://pith.science/pith/V4D267W4SOMKDMREQVRCLBVOWS.json","graph_json":"https://pith.science/api/pith-number/V4D267W4SOMKDMREQVRCLBVOWS/graph.json","events_json":"https://pith.science/api/pith-number/V4D267W4SOMKDMREQVRCLBVOWS/events.json","paper":"https://pith.science/paper/V4D267W4"},"agent_actions":{"view_html":"https://pith.science/pith/V4D267W4SOMKDMREQVRCLBVOWS","download_json":"https://pith.science/pith/V4D267W4SOMKDMREQVRCLBVOWS.json","view_paper":"https://pith.science/paper/V4D267W4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2304.01982&json=true","fetch_graph":"https://pith.science/api/pith-number/V4D267W4SOMKDMREQVRCLBVOWS/graph.json","fetch_events":"https://pith.science/api/pith-number/V4D267W4SOMKDMREQVRCLBVOWS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/V4D267W4SOMKDMREQVRCLBVOWS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/V4D267W4SOMKDMREQVRCLBVOWS/action/storage_attestation","attest_author":"https://pith.science/pith/V4D267W4SOMKDMREQVRCLBVOWS/action/author_attestation","sign_citation":"https://pith.science/pith/V4D267W4SOMKDMREQVRCLBVOWS/action/citation_signature","submit_replication":"https://pith.science/pith/V4D267W4SOMKDMREQVRCLBVOWS/action/replication_record"}},"created_at":"2026-07-05T08:05:48.408229+00:00","updated_at":"2026-07-05T08:05:48.408229+00:00"}